The 2011 Seventh International Conference on Intelligent...
Transcript of The 2011 Seventh International Conference on Intelligent...
1
The 2011 Seventh International Conference
on Intelligent Computing
2011 Sino-Korean Intelligent Information
Processing and Automation Workshop
August 11-14, 2011, Zhengzhou, China
2
FINAL PROGRAM
The 2011 Seventh International Conference on Intelligent Computing
2011 Sino-Korean Intelligent Information
Processing and Automation Workshop
August 11-14, 2011 Zhengzhou, Henan, China
3
Outlines
Welcome Message 4
ICIC2011 Organization 5
Program Committee Members 5
Reviewers 7
Sponsors 9
The Location of Conference Venue in Zhengzhou 10
The 4th Floor of VIP Wing of FengLeYuan Hotel 11
Schedule Overview 12
Introduction of Keynote Speakers 13
Parallel Sessions for Oral Presentations 19
Abstract of Papers 21
4
WELCOME FROM THE ORGANIZERS
The International Conference on Intelligent Computing (ICIC) was formed to provide an annual forum dedicated to the emerging and challenging topics in artificial intelligence, machine learning, bioinformatics, and computational biology, etc. It aims to bring together researchers and practitioners from both academia and industry to share ideas, problems and solutions related to the multifaceted aspects of intelligent computing.
ICIC2011, held in Zhengzhou, China, August 11-14, 2011, constituted the Seventh International Conference on Intelligent Computing. It built upon the success of ICIC2010, ICIC2009, ICIC2008, ICIC2007, ICIC 2006, and ICIC 2005 held in Changsha/China, Ulsan/Korea, Shanghai, Qingdao, Kunming and Hefei, China, respectively.
This year, the conference concentrated mainly on the theories and methodologies as well as the emerging applications of intelligent computing. Its aim was to unify the picture of contemporary intelligent computing techniques as an integral concept that highlights the trends in advanced computational intelligence and bridges theoretical research with applications. Therefore, the theme for this conference was “Advanced Intelligent Computing Technology and Applications”. Papers focused on this theme were solicited, addressing theories, methodologies, and applications in science and technology.
ICIC 2011 received 832 submissions from 28 countries and regions. All papers went through a rigorous peer review procedure and each paper received at least three review reports. Based on the review reports, the Program Committee finally selected 281 high-quality papers for presentation at ICIC 2011, which are included in three volumes of proceedings published by Springer: one volume of Lecture Notes in Computer Science (LNCS) 6838, one volume of Lecture Notes in Artificial Intelligence (LNAI) 6839, and one volume of Lecture Notes in Bioinformatics (LNBI) 6840. In addition, among them, the 10 and 44 high-quality papers have also respectively been recommended to BMC Bioinformatics and Neurocomputing for possible publication.
The organizers of ICIC 2011, Zhengzhou University of Light Industry, and Hefei Institute of Intelligent Machines of Chinese Academy of Sciences made an enormous effort to ensure the success of ICIC 2011. We hereby would like to thank the members of the Program Committee and the referees for their collective effort in reviewing and soliciting the papers. We would like to thank Alfred Hofmann, executive editor from Springer, for his frank and helpful advice and guidance throughout and for his continuous support in publishing the proceedings. In particular, we would like to thank all the authors for contributing their papers. Without the high-quality submissions from the authors, the success of the conference would not have been possible. Finally, we are especially grateful to the IEEE Computational Intelligence Society, the International Neural Network Society, and the National Science Foundation of China for their sponsorship.
August, 2011
De-Shuang Huang
DeLiang Wang YanLi Lv
5
ICIC 2011 Organization
General Co-Chairs:
De-Shuang Huang, China DeLiang Wang, USA Yanli Lv, China
Program Committee Co-Chairs: Zhongming Zhao, USA Kang-Hyun Jo, Korea Jianhua Ma, Japan
Organizing Committee Co-Chairs: Yong Gan, China Sushi Zhang, China Hong-Qiang Wang, China Wei Jia, China
Award Committee Chair Laurent Heutte, France
Publication Chair: Juan Carlos Figueroa, Colombia
Special Session Chair: Phalguni Gupta, India
Tutorial Chair: Vitoantonio Bevilacqua, Italy
International Liaison Chair: Prashan Premaratne, Australia
Publicity Co-Chairs Xiang Zhang, USA Kyungsook Han, Korea Lei Zhang, Hong Kong, China
Exhibition Chair Xueling Li, China
Organizing Committee Members Conference Secretary:
Xunlin Zhu, China Shengli Song, China Haodong Zhu, China Xiaoke Su, China Xueling Li, China Jie Gui, China Zhi-Yang Chen, China
Program Committee Members
Andrea Francesco Abate, Italy Vasily Aristarkhov, Russian Federation Costin Badica, Romania Shuhui Bi, Japan David B. Bracewell, USA Martin Brown, UK Zhiming Cai, Macau, China Chin-chih Chang, Taiwan,
China Pei-Chann Chang, China Guanling Chen, USA Jack Chen, Canada Shih-Hsin Chen, China Wen-Sheng Chen, China Xiyuan Chen, China Yang Chen, China Yuehui Chen, China Ziping Chiang, China
Michal Choras, Poland Angelo Ciaramella, Italy Jose Alfredo F. Costa, Brazil Youping Deng, USA Eng. Salvatore Distefano, Italy Mariagrazia Dotoli, Italy Meng Joo Er, Singapore Ahmed Fadiel, USA Karim Faez, Iran
6
Jianbo Fan, China Minrui Fei, China Wai-Keung Fung, Canada Jun-Ying Gan, China Liang Gao, China Xiao-Zhi Gao, Finland Carlos Alberto Reyes Garcia, Mexico Dunwei Gong, China Valeriya Gribova, Russia M. Michael Gromiha, Japan Kayhan Gulez, Turkey Anyuan Guo, China Phalguni Gupta, India Sung Ho Ha, Korea Fei Han, China Kyungsook Han, Korea Nojeong Heo, Korea Laurent Heutte, France Wei-Chiang Hong, Taiwan, China Zeng-Guang Hou, China Yuexian Hou, China Kun Huang, USA Peter Hung, Ireland Sajid Hussain, USA Peilin Jia, USA Minghui Jiang, China Zhenran Jiang, China Kang-Hyun Jo, Korea Yoshiaki Kakuda, Japan Sanggil Kang, Korea Muhammad Khurram Khan, Saudi Arabia Sungshin Kim, Korea In-Soo Koo, Korea Bora Kumova, Turkey Yoshinori Kuno, Japan Wen-Chung Kuo, Taiwan, China Takashi Kuremoto, Japan
Vincent C S Lee, Australia Guo-Zheng Li, China Jing Li, USA Kang Li, UK Peihua Li, China Ruidong Li, Japan Shutao Li, China Xiaoou Li, Mexico Hualou Liang, USA Honghuang Lin, USA Chunmei Liu, USA Liu Chun-Yu Liu, USA Ju Liu, China Van-Tsai Liu, Taiwan, China Jinwen Ma, China Tarik Veli Mumcu, Turkey Igor V. Maslov, Japan Filippo Menolascina, Italy Primiano Di Nauta, Italy Roman Neruda, Czech Republic Ben Niu, China Sim-Heng Ong, Singapore Ali Özen, Turkey Vincenzo Pacelli, Italy Francesco Pappalardo, Italy Witold Pedrycz, Canada Caroline Petitjean, France Pedro Melo-Pinto, Portugal Susanna Pirttikangas, Finland Prashan Premaratne, Australia Daowen Qiu, China Yuhua Qian, China Seeja K R, India Marylyn Ritchie, USA Ivan Vladimir Meza Ruiz, Mexico Fariba Salehi, Iran Angel Sappa, Spain
Jiatao Song, China Stefano Squartini, Italy Hao Tang, China Antonio E. Uva, Italy Jun Wan, USA Bing Wang, USA Ling Wang, China Xue Wang, China Xuesong Wang, China Yong Wang, Japan Yufeng Wang, Japan Zhong Wang, USA Wei Wei, Norway Zhi Wei, China Ling-Yun Wu, China Junfeng Xia, USA Shunren Xia, China Hua Xu, USA Jianhua Xu, China Shao Xu, Singapore Ching-Nung Yang, Taiwan, China Wen Yu, Mexico Zhi-Gang Zeng, China Jun Zhang, China Xiang Zhang, USA Yanqing Zhang, USA Zhaolei Zhang, Canada Lei Zhang, Hong Kong, China Xing-Ming Zhao, China Zhongming Zhao, USA Chun-Hou Zheng, China Huiru Zheng, UK Bo-Jin Zheng, China Fengfeng Zhou, USA Mianlai Zhou, China Li Zhuo, China Yuhua Qian, China
7
Reviewers:
Ibrahim SAHIN, Bora Kumova, Birol Soysal, Yang Xiang, Gang Feng, Francesco Camastra, Antonino Staiano, Alessio Ferone, Surya Prakash, Badrinath Srinivas, Dakshina Ranjan Kisku, Zilu Ying, Guohui HE, Vincenzo Pacelli, Pasqualedi Biase, federicamiglietta, junying zeng, yibin yu, Kaili Zhou, Yikui Zhai, WenQiang Yang, WenJu Zhou, Dae-Nyeon Kim, Ilmari Juutilainen, Alessandro Cincotti, Marzio Alfio Pennisi, Carme Julià, Santo Motta, Nestor Arana-Arexolaleiba, Myriam Delgado, Giuliana Rotunno, Agostino Marcello Mangini, Carson K.Leung, Gabriella Stecco, Yaser Maddahi, Jun Wan, Jiajun Bracewell, Jing Huang, Kunikazu Kobayashi, Feng Liangbing, JoaquinTorres-Sospedra, Takashi Kuremoto, Fabio Sciancalepore, Valentina Boschian, Chuang Ma, juanxiao, Lihua Jiang, Changan Jiang, NiBu, Shengjun Wen, Aihui Wang, Peng Wang, Myriam Delgado, Wei Ding, Kurosh Zarei-nia, Li Zhu, Hoang-HonTrinh, Alessia Albanese, Song Zhu, Lei Liu, fengjiang, Bo Liu, Ye Xu, Gang Zhou, ShengyaoWang, Yehu Shen, Liya Ding, Hongjun Jia, Hong Fu, Tiantai GUO, Liangxu Liu, dawen xu, Zhongjie Zhu, JayasuhaJ.S., Aravindan Chandrabose, shanthi KJ, Shih-Hsin Chen, Wei-Hsiu Huang, Antonio Maratea, Sandra Venske, Carolina Almeida, Richard Goncalves, minggao, Feng Li, Yu Xue, Qin Ma, MING GAO, Gang Xu, Yandong Zhang, Benhuai Xie, Ran Zhang, MingkunLi, Zhide Fang, Xiaodong Yang, Lein Harn, Wu-Chuan Yang, Bin Qian, Quan-ke PAN, Junqing Li, qiaowei, Xinli Xu, Hongjun Song, Michael Gromiha, Xueling Li, Y-h.Taguchi, Yu-Yen Ou, Hong-Bin Shen, Ximo Torres, weidong yang, Quanming Zhao, ChongShen, Xianfeng Rui, Phalguni Gupta, Yuan Xu, yuefang zhao, custianacucu, wangxiaojuan, Guihui Zhang, Xinyu LI, Yang Shi, Hongcheng Liu, Lijun Xu, Xiaomin Liu, Tonghua Su, Junbiao Pang, chunnie, Saihua Lin, Alfredo Pulvirenti, PedroMelo-Pinto, Armando Fernandes, Atsushi Yamashita, Kazunori Onoguchi, Liping Zhang, Qiong Zhu, chizhou, qirong mao, Lingling Wang, WenYong Dong, wenwen shen, Gang Bao, Shiping Wen, giorgioiacobellis, Paolo Lino, Qi Jiang, Yan-JieLi, Gurkan Tuna, Tomoyuki Ohta, Jianfei Hu, Xueping Yu, ShinjiInoue, Eitaro Kohno, Rui-Wei Zhao, Shixing Yan, Jiaming Liu, Wen-ChungKuo, Jukka Riekki, Jinhu Lu, Qinglai Wei, Michele Scarpiniti, Simone Bassis, Zhigang Liu, Pei Wang, qianyufeng, Jingyi Qu, Mario Foglia, michelefiorentino, Luciano Lamberti, Lein Harn, Kai Ye, Zhenyu Xuan, Francesco Napolitano, RaphaelIsokpehi, Vincent Agboto, Ryan Delahanty, Shaohui Liu, Ching-Jung Ting, Chuan-Kang Ting, Chien-lung Chan, Jyun-Jie Lin, Liang-Chih Yu, Richard Tzong-Han Tsai, Chin-Sheng Yang, Jheng-LongWu, Jun-Lin Lin, Chia-Yu Hsu, Wen-Jia Kuo, Yi-Kuei Lin, K.Robert Lai, Sumedha Gunewardena, Qian Xiang, Joe Song, Ryuzo Okada, Handel Cheng, Chin-Huang Sun, Tung-Chen Huang, Bin Yang, Changyan Xiao, Mingkui Tan, zhigangling, leizhou, Hung-Chi Su, Chyuan-Huei Yang, Rey-Sern Lin, Cheng-Hsiung Chiang, Chrisil Arackaparambil, Valerio Bianchi, Zhi Xie, Ka-Chun Wong, Zhou Yong, Aimin Zhou, YONG ZHANG, Yan Zhang, jihui zhang, Xiangjuan Yao, Jing Sun, jianyongsun, Yi-nan Guo, yongbinzhang, Vasily Aristarkhov, Hongyan SANG, Aboubekeur Hamdi-Cherif, chenbo, minli, Linlin Shen, Jianwei Yang, Lihua Guo, Manikandan Narayanan, MasoumehEsfandiari, Amin Yazdanpanah, Ran Tao, Weiming Yu, Aditya Nigam, Kamlesh Tiwari, MariaDe Marsico, StefanoR, Wei Wei, Lvzhou Li, HaozhenSitu, Bian Wu, linhuazhou, Shaojing Fan, Qingfeng Li, Rina Su, Hongjun Song, Bin Ye, zhaojun, Yindi ZHAO, Kun Tan, chenwei, Yuequan Yang, qianzhang, Zhigang Yan, Jianhua Xu, Ju-Yin Cheng, Yu Gu, Guang Zeng, Xuezheng Liu, yuanweirong, Ren Xinjun, yufutian, Mingjing Yang, Chunjiang Zhang, Yinzhi Zhou, William Carswell, Andrey Vavilin, Sang-Hee Lee, Yan Fan, Hong Wang, Fangmin Yao, Angelo Ciaramella,
8
Eric Hsu, Xiao-Feng Wang, Jing Deng, Wanqing Zhao, weihuadeng, Xueqin Liu, SungShin Kim, Gyeongdong Baek, Seongpyo Cheon, bilalkhan, Maqsood Mahmud, Pei-Wei Tsai, Lin Zhang, Bo Peng, Jifeng Ning, Yongsheng Dong, Chonglun Fang, Yan Yang, Hongyan Wang, Min Wang, Rong-Xiang Hu, Xiaoguang Li, jingzhang, Yue Jiao, Hui Jing, Ruidong Li, Wei Xiong, Toshiaki Kondo, Suresh Sundaram, haimin, Donghui Hu, Xiaobin Tan, Stefano Dell'Atti, Rafal Kozik, Michal Choras, RPhan, Yuan-Fang Li, Tsung-Che Chiang, Ming Xia, Weimin Huang, Xinguo Yu, Sabooh Ajaz, ZhengMao Zou, Prashan Premaratne, Ibrahim ALISKAN, YUSUFALTUN, Ali Ahmed Adam, Janset Dasdemir, Turker Turker, Ibrahim Kucukdemiral, JunSheng Zhou, Yue Wang, Yoshiaki Kakuda, Daqiang Zhang, Min-Chih Chen, Aimin Zhou, Shihong Ding, Ziping Chiang, Xiaoyu Wen, Gao Liang, Orion Reyes-Galaviz, Miguel Mora-Gonzalez, Pilar Gomez-Gil, Miguel Mora-Gonzalez, Jida HUANG, Insookoo, Nhan Nguyen-Thanh, ThucKieu Xuan, Yang Zhao, Andreas Konstantinidis, Canyi Lu, Nobuo Funabiki, Yukikazu Nakamoto, Xin Zhou, Qian Wang, Xiaoyan Yin, JUANCUI, Francesco Polese, senjia, Crescenzio Gallo, yusun, Xuewen Xia, chuanpeng, Chen Jing-Yuan, Edison Yu, Petra Vidnerová, Klara Peskova, Martin Pilat, Liu Zhaochen, Jun Du, Ning Lv, Yoko Kamidoi, Meng Wang, Hao Xin, Dingfei Ge, Xin Gao, Ivan;Vladimir Meza;Ruiz, Tsang-Yi Wang, Sangyoon Oh, Li Ruichang, Fan Jing, Lin Wang, Chunlu Lai, hamidecheraghchi, Wen-Tsai Sung, TheAnh BUI, Zhong Qishui, Duyu Liu, keliangjun, Ying Qiu, Huisen Wang, Maria Elena Valcher, Alex Muscar, SorinIlie, Amelia Badica, liuguanghai, Changbin Du, Jianqing Li, Hao Wang, Yurong Cheng, Mingyi Wang, Claudio Franca, Jose Alfredo Ferreira Costa, Tomasz Andrysiak, Ajay Kumar, Lei Zhang, Zhoumian Wang, Ji-Xiang Du, Xibei Yang, junhong wang, Wei Wei, guoping Lin, Dun Liu, changzhongwang, Xiaoxiao Ma, xiao xueyang, Wei Yu, Ming Yang, Francesca Nardone, Kok-Leong Ong, David Taniar, Nali Zhu, Hailei Zhang, My HaLe, Haozhen Situ, Lvzhou LI, Mianlai Zhou, CHIN-CHIH CHANG, Carlos;A.Reyes-Garcia, Jack Chen, Wankou Yang, Qijun Zhao, jinxie, Xian Chen, Gustavo Fontoura, Xiaoling Zhang, Ondrej Kazik, Bo Yan, ZZ, Yun Zhu, B.Y.Lee, Jianwen Hu, Keling Chang, Jianbo Fan, Chunming Tang, Hongwei Ma, Valeria Gribova, Valeria Gribova, Ailong Wu, William-Chandra Tjhi, Gongqing Wu, yaohongliang, Bingjing Cai, Lin Zhu, Li Shang, Bo Li, Jun Zhang, Peng Chen, wenlongsun, xiaoli Wei, Bing Wang, Jun Zhang, Peng Chen, karimfaez, Xiaoyan Wang, Wei-Chiang Hong, Chien-Yuan Lai, Sugang Xu, Junfeng Xia, Yi Xiong, Xuanfang Fei, Jingyan Wang, Zhongming Zhao, Yonghui Wu, Samir Abdelrahman, Mei Liu, Fusheng Wang, Shao-Lun Lee, Wen zhang, Zhi-Ping Liu, Qiang Huang, Jiguang Wang, Rui Xue, Xiao Wang, jibinqu, Bojin Zheng, Susanna Pirttikangas, ukasz Saganowski, Chunhou Zheng, Zheng Chunho, meijun, Geir Solskinnsbakk, Satu Tamminen, Laurent HEUTTE, Mikko Perttunen, Renqiang Min, RongGui Wang, xinping xie, horace wang, Hong-Jie Yu, Wei Jia, Wang huqing.
9
Sponsors Hosted by
Zhengzhou University of Light Industry,
Zhengzhou, China
Hefei Institute of Intelligent Machines, Chinese
Academy of Sciences
University of Science & Technology of China
Chinese Academy of Sciences
Financially Supported by
The National Science Foundation of China
Technically co-sponsored by
The International Neural Network Society
The IEEE Computational Intelligence Society
10
The Location of Conference Venue in Zhengzhou
Conference Venue
ICIC2011 Conference Venue is Henan Feng Le Yuan (FLY) hotel, which is a five-four-three star hybrid tourist hotel located in North Nanyang Road, a key economic backbone of the northbound section of Zhengzhou City, with convenient and smoothly traffic, it is the ideal abode hotel for both business and tourism. The hotel can offer west and Chinese food and beverages, with six halls and 50 luxury banquet hall and bar, has a capacity of 2,200 meals at the same time, it has most seats of all the hotels in Zhengzhou City. The FLY hotel was named the "most trend hotel," it owns 300 sets of different characteristics of the senior rooms, deluxe rooms, suites and standard distinguished presidential suite. Conference Centre possession of small and large conference rooms of 12, with advanced facilities, accommodates various conferences and business activities. The tropical rain forest hot springs spa museum is new project, the full-function and high grades are the most in Asia, opened Spa leisure precedent in Henan.
FengLeYuan Hotel
12
Schedule Overview
Time Date Morning Afternoon Evening
Aug. 11
Thursday
Registration (14:30-20:30pm)
Location: Lobby of VIP Wing of FengLeYuan Hotel
Aug. 12
Friday
Opening Ceremony Session
Location: Hall on the 4th floor of the
FengLeYuan Hotel
8:30-9:15am
Plenary Speech
Location: Hall on the 4th floor of the
FengLeYuan Hotel
Chair: De-Shuang Huang
Speaker: Gary G. Yen
9:15-9:50 am
Coffee Break, 9:50-10:00am
Chair: De-Liang Wang
Speaker: Marios Polycarpou
10:00-10:35am
Chair: Kang-Hyun Jo
Speaker: Vincenzo Piuri
10:40-11:15am
Chair: Phalguni Gupta
Speaker: Donald C. Wunsch II
11:20-11:55
Lunch
Oral presentation
13:40-15:40pm
Room A, B, C
Coffee Break: 15:40-15:50pm
Oral presentation
15:50-17:50pm
Room A, B, C
Welcome Reception
Hall on the ground floor of VIP Wing of FengLeYuan Hotel 18:30-20:00pm
Aug. 13
Saturday
Oral presentation
8:00-10:00am
Room A, B, C
Coffee Break: 10:00-10:10am
Oral presentation
10:10-12:10am
Room A, B, C
Lunch
Oral presentation
13:40-15:40pm
Room A, B, C
Coffee Break: 15:40-15:50pm
Oral presentation
15:50-17:50pm
Room A, B, C
Technical Committee Meeting
14:30 – 18:00pm
Banquet
To be announced 18:30-21:00pm
Aug. 14
Sunday Tour to Shaolin Temple, etc
Dinner 18:30-20:00
13
ICIC2011 Keynote Speakers
Keynot Speakers 1: Gary G. Yen
Cultual-Based Particle Swarm Optimization for Constrained Multiobjective Optimization
Gary G. Yen, Professor, Ph.D., IEEE Fellow
School of Electrical and Computer Engineering, Oklahoma State University, USA
Abstract: Evolutionary computation is the study of biologically motivated computational paradigms which exert novel ideas and inspiration from natural evolution and adaptation. The applications of population-based heuristics in solving constrained and dynamic optimization problems have been receiving a growing interest from computational intelligence community. Most practical optimization problems are with the existence of constraints and uncertainties in which the fitness function changes through time and is subject to multiple constraints. In this study, we propose the cultural-based particle swarm optimization (PSO) to solve these problems with real-world complications. A cultural framework is introduced that incorporates the required information from the PSO into five sections of the belief space, namely situational knowledge, temporal knowledge, domain knowledge, normative knowledge, and spatial knowledge. The archived information is exploited to detect the changes in the environment and assists response to the change and constraints through a diversity based repulsion among particles and migration among swarms in the population space, also helps in selecting the leading particles in three different levels, personal, swarm, and global level. Comparison of the proposed cultural based PSO over numerous challenging constrained and dynamic benchmark problems demonstrates the competitive, if not appreciably much better, performance with respect to selected state-of-the-art PSO heuristics.
Biography: Gary G. Yen received the Ph.D. degree in electrical and computer engineering from the University of Notre Dame, Notre Dame, Indiana in 1992. He is currently a Professor in the School of Electrical and Computer Engineering, Oklahoma State University. Before he joined OSU in 1997, he was with the Structure Control Division, U.S. Air Force Research Laboratory in Albuquerque, NM. His research is supported by the DoD, DoE, EPA, NASA, NSF, and Process Industry. His research interest includes intelligent control, computational intelligence,
evolutionary multiobjective optimization, conditional health monitoring, signal processing and their industrial/defense applications.
Gary was an associate editor of the IEEE Transactions on Neural Networks and IEEE Control Systems Magazine during 1994-1999, and of the IEEE Transactions on Control
14
Systems Technology, IEEE Transactions on Systems, Man and Cybernetics and IFAC Journal on Automatica and Mechatronics during 1999-2006. He is currently serving as an associate editor for the IEEE Transactions on Evolutionary Computation. He served as the General Chair for the 2003 IEEE International Symposium on Intelligent Control held in Houston, TX and 2006 IEEE World Congress on Computational Intelligence held in Vancouver, Canada. Gary served as Vice President for the Technical Activities, IEEE Computational Intelligence Society in 2004-2005 and is the founding editor-in-chief of the IEEE Computational Intelligence Magazine from 2006 to 2009. He is currently serving as President of the IEEE Computational Intelligence Society in 2010-2011. He is a Fellow of IEEE.
Keynot Speakers 2: Marios Polycarpou
Intelligent Distributed Fault Diagnosis of Interconnected Dynamical Systems
Marios M. Polycarpou, Professor, Ph.D, IEEE Fellow
Department of Electrical and Computer Engineering, University of Cyprus, Cyprus
Abstract: Electronic devices are starting to become widely available for monitoring and controlling large-scale distributed systems. These devices may include sensing capabilities for on-line measurement, actuators for controlling certain variables, microprocessors for processing information and making real-time decisions based on designed algorithms, and telecommunication units for exchanging information with other electronic devices or possibly with human operators. A collection of such devices may be referred to as a networked intelligent agent system. Such systems have the capability to generate a huge volume of spatial-temporal data that can be used for monitoring and control applications of large-scale distributed systems. One of the most important research challenges in the years ahead is the development of information processing methodologies that can be used to extract meaning and knowledge out of the ever-increasing electronic information that will become available. Even more important is the capability to utilize the information that is being produced to design software and devices that operate seamlessly, autonomously and reliably in some intelligent manner. The ultimate objective is to design networked intelligent agent systems that can make appropriate real-time decisions in the management of large-scale distributed systems, while also providing useful high-level information to human operators.
One of the most important classes of large-scale distributed systems deals with the reliable operation and intelligent management of critical infrastructures, such as electric power systems, telecommunication networks, water systems, and transportation systems. The design, control and fault monitoring of critical infrastructure systems is becoming increasingly more challenging as their size, complexity and interactions are steadily growing. Moreover, these critical infrastructures are susceptible to natural disasters, frequent failures, as well as malicious attacks. There is a need to develop a common system-theoretic fault diagnostic framework for critical infrastructure systems and to
15
design architectures and algorithms for intelligent monitoring, control and security of such systems. The goal of this presentation is to motivate the need for health monitoring, fault diagnosis and security of critical infrastructure systems and to provide a fault diagnosis methodology for detecting, isolating and accommodating both abrupt and incipient faults in a class of complex nonlinear dynamic systems. A detection and approximation estimator based on computational intelligence techniques is used for online health monitoring. Various adaptive approximation techniques and learning algorithms will be presented and illustrated, and directions for future research will be discussed.
Biography: Marios M. Polycarpou is a Professor of Electrical and Computer Engineering and the Director of the KIOS Research Center for Intelligent Systems and Networks at the University of Cyprus. He received the B.A. degree in Computer Science and the B.Sc. degree in Electrical Engineering both from Rice University, Houston, TX, USA in 1987, and the M.S. and Ph.D. degrees in Electrical Engineering from the University of Southern California, Los Angeles, CA, in 1989 and 1992 respectively. In 1992, he joined the University of Cincinnati, Ohio, USA, where he reached
the rank of Professor of Electrical and Computer Engineering and Computer Science. In 2001, he was the first faculty to join the newly established Department of Electrical and Computer Engineering at the University of Cyprus, where he served as founding Department Chair from 2001 to 2008. His teaching and research interests are in intelligent systems and control, adaptive and cooperative control systems, computational intelligence, fault diagnosis and distributed agents. Dr. Polycarpou has published more than 200 articles in refereed journals, edited books and refereed conference proceedings, and co-authored the book Adaptive Approximation Based Control, published by Wiley in 2006. He is also the holder of 3 patents.
Prof. Polycarpou has served as the Editor-in-Chief of the IEEE Transactions on Neural Networks between 2004-2010. He serves as an Associate Editor of two international journals and is past Associate Editor of the IEEE Transactions on Neural Networks (1998-2003) and of the IEEE Transactions on Automatic Control (1999-2002). He served as the Chair of the Technical Committee on Intelligent Control, IEEE Control Systems Society (2003-05) and as Vice President, Conferences, of the IEEE Computational Intelligence Society (2002-03). He is currently an elected member of the Board of Governors of the IEEE Control Systems Society, an elected AdCom member of the IEEE Computational Intelligence Society, and the Chair of Awards Committee for the IEEE Computational Intelligence Society. Dr. Polycarpou was the recipient of the William H. Middendorf Research Excellence Award at the University of Cincinnati (1997) and was nominated by students for the Professor of the Year award (1996). He has been invited as Keynote Plenary Speaker at 16 international conferences during the last five years and is currently an IEEE Distinguished Lecturer in computational intelligence. He participated in more than 50 research projects/grants, funded by several agencies and industry in the United States, by the European Commission and by the
16
Research Promotion Foundation of Cyprus. Dr. Polycarpou is a Fellow of the IEEE and the President-Elect of the IEEE Computational Intelligence Society.
Keynot Speakers 3: Vincenzo Piuri
Neural Techniques for 3D Surface Reconstruction
Vincenzo Piuri, Professor, Ph.D, IEEE Fellow
Università degli Studi di Milano, Italy
Abstract: Applications based on three-dimensional object models are today very common, and can be found in many fields as design, archeology, medicine, and entertainment. A digital 3D model can be obtained, for example, by means of physical object measurements performed by using a 3D scanner. In this approach, an important step of the 3D model building process consists of creating the object's surface representation from a cloud of noisy points sampled on the object itself. This process can be viewed as the estimation of a function from a finite subset of its points.
Problems of this kind occur in many branches of applied mathematics, and computer science. Many techniques have been developed to face them, such as interpolation, extrapolation, regression analysis, and curve fitting. In computational intelligence this problem is viewed as a supervised learning problem, where the two-dimensional vector coordinates of the single point is an input instance, while the third coordinate is considered as an output label. The approximation function identifies how to obtain labels from instances. Several effective computational intelligence paradigms have been developed for solving these kinds of problems. For the solution of the function reconstruction problem, neural techniques, generally, show a good trade-off between computational complexity, accuracy and robustness of the solution with respect to other methods. In this context, there are many different paradigms which are able to find the approximation function, e.g., Multi-layer Perceptron Networks, Radial Basis Function (RBF) Networks, and Support Vector Machines (SVM). In general, there is not a single paradigm better than the others, but each one performs differently depending on the application context. This keynote speech is directed to introduce the fundamental needs of the applications mentioned above, to briefly overview the techniques for surface reconstruction, to analyze and discuss in detailed the neural techniques suited for addressing this problem, and to present the most recent results of research.
Biography: Vincenzo Piuri obtained the Ph.D. in Computer Engineering in 1989, at Politecnico di Milano, Italy. From 1992 to September 2000, he was Associate Professor in Operating Systems at Politecnico di Milano. Since October 2000 he is Full Professor in Computer Engineering at the Università degli Studi di Milano, Italy. He was Visiting Professor at the University of Texas at Austin during the summers from 1993 to 1999. His
17
research interests include: theory and industrial applications of neural networks, intelligent measurement systems, signal and image processing for industrial applications, biometrics, application-specific processing architectures, digital signal processing architectures, and fault tolerance. Original results have been published in more than 300 papers in book chapters, international journals, and proceedings of international conferences.
He is Fellow of the IEEE, Distinguished Scientist of ACM, and Senior Member of the INNS. He was Associate Editor of the IEEE Transactions on Neural Networks and the IEEE Transactions on Instrumentation and Measurement. He was President of the IEEE Computational Intelligence Society (2006-07), Vice President for Publications both of the IEEE Instrumentation and Measurement Society and the IEEE Systems Council, Vice President for Members Activities of the IEEE Neural Networks Society, Vice President for Education of the IEEE Biometrics Council, and Member of the Administrative Committee both of the IEEE Instrumentation and Measurement Society and the IEEE Computational Intelligence Society. He is IEEE Division X Director (2010-2012). In 2002 he received the IEEE Instrumentation and Measurement Society Technical Award for his contributions to the advancement of computational intelligence theory and practice in measurement systems and industrial applications. More information are available at http://www.dti.unimi.it/piuri.
Keynot Speakers 4: Donald C. Wunsch II
Hierarchical Clustering
Donald C. Wunsch II, Ph.D. Professor, IEEE Fellow, INNS Senior Fellow,
Missouri University of Science & Technology, USA
http://people.mst.edu/faculty/dwunsch_profile.html
Biography: Donald Wunsch is the M.K. Finley Missouri Distinguished Professor at Missouri University of Science & Technology (Missouri S&T). Earlier employers were Texas Tech University, Boeing, Rockwell International, and International Laser Systems. His education includes Executive MBA - Washington University in St. Louis, Ph.D., Electrical Engineering - University of Washington (Seattle), M.S., Applied Mathematics (same institution), B.S., Applied Mathematics - University of New Mexico. Key research
contributions are Clustering; Adaptive resonance and Reinforcement Learning architectures, hardware and applications; Neurofuzzy regression; Traveling Salesman Problem heuristics; Robotic Swarms; and Bioinformatics. He has produced 15 Ph.D. recipients in Computer Engineering, Electrical Engineering, and Computer Science; has attracted over $8 million in sponsored research; and has over 300 publications including nine books. His research has been cited well over
18
1900 times, excluding self-citations of all authors. He is an IEEE Fellow and previous INNS President, INNS Fellow and Senior Fellow 07 – present, and served as IJCNN General Chair, and on several Boards, including the St. Patrick’s School Board, IEEE Neural Networks Council, International Neural Networks Society, and the University of Missouri Bioinformatics Consortium. He chairs the Missouri S&T Information Technology and Computing Committee, a Faculty Senate Standing Committee.
19
Parallel Sessions for Oral Presentations
Room Time
Room A Room B Room C
Afternoon Aug. 12 13:40-15:40
Machine Learning Theory and Methods I Chair: Wangmeng Zuo Paper No: 1650,1579, 1901, 1637, 1746,1768, 1856, 1679
Neural Networks Chair: Lingzhi Wang Paper No: 1375, 1463, 1576, 1416 , 1437,1183, 1260,
Intelligent Computing in Bioinformatics & Bio-medical Engineering Chair: M. Michael Gromiha & Jingyan WangPaper No: 1987, 1690, 1621, 1984, 1494, 1294, 1243, 1983,
Afternoon Aug. 12 15:50-17:50
Machine Learning Theory and Methods II Chair: Rudy Rotili Vincenzo Di Lecce Paper No: 1484, 1510, 1172, 1036, 1653, 1436, 1442, 1473,
Intelligent Control and Automation Chair: Hongyan Yan Paper No: 1219, 1869, 1918, 1449, 1749, 1939, 1309, 1415, 1196
Intelligent Computing in Pattern Recognition Chair: Linlin Shen Paper No: 1591, 1541, 1194, 1483, 1694,1860,1657, 1411
Morning Aug. 13 8:00-10:00
Evolutionary Learning & Genetic Algorithms & Optimization Chair: Roman Neruda Syeda Darakhshan Jabeen Paper No: 1563, 1453, 1007, 1079, 1835 ,1585, 1905,1663, 1906
Intelligent Computing Theory and Application I Chair: Shulin Wang Xutang Zhang Paper No: 1920, 1996, 1678, 1995, 1244, 1998, 1070, 1590, 1433
Sino-Korean Intelligent Information Processing and Automation Workshop I Chair: De-Shuang Huang
Kang-Hyun Jo Paper No: 1156, 1588, 1866, 1256, 1124, 1126, 1241, 1325, 1400
Morning Aug. 13 10:10-12:10
Swarm Intelligence and Optimization Chair: Mehran Yazdi
Yanxia Sun Paper No: 1624, 1258, 1302,1575, 1689, 1565, 1348, 1273, 1865
Intelligent Computing Theory and Application II Chair: Xibei Yang Paper No: 1493, 1855, 1444, 1445, 1474, 1486, 1098, 1516, 1763, 1245,
Sino-Korean Intelligent Information Processing and Automation Workshop II Chair: Kang-Hyun Jo
De-shuang Huang Paper No: 1367, 1413, 1618, 1649, 1475, 1491, 1566, 1644, 1770
20
Room Time
Room A Room B Room C
Afternoon Aug. 13 13:40-15:40
Intelligent Computing in Computer Vision Chair: Tarik Veli Mumcu Paper No: 1129, 1489, 1418, 1439, 1582, 1646, 1540, 1571
Sino-Korean Intelligent Information Processing and Automation Workshop III Chair: Hee-Jun Kang Hong-Hee Lee Paper No: 1069, 1861, 1306, 1917, 1772, 1854, 1902, 1911, 1922
Afternoon Aug. 13 15:50-17:50
Intelligent Computing in Image Processing Chair: Vandana Dixit Kaushik
Satoshi Mori Paper No: 1654, 1669, 1698, 1296, 1564, 1655, 1305, 1308,
Workshop on Intelligent Computing in Scheduling Chair: Ling Wang Paper No: 1078, 1151,1155, 1099 , 1094, 1215, 1556, 1395, 1832, 1587, 1515, 1220, 1195,1259 Technical Committee Meeting Chair: De-Shuang Huang
Sino-Korean Intelligent Information Processing and Automation Workshop IV Chair: Hong-Hee Lee Myung Jae Yi Paper No: 1592, 1535, 1645,1236, 1247, 1326,1324
21
Abstract of Papers
Afternoon, Friday, August 12
Room A Machine Learning Theory and Methods I
Chair Wangmeng Zuo 13:40 – 13:55 Afternoon Aug. 12 Room A Paper No: 1650
Importance Weighted AdaRank
Shangkun Ren1, Yuexian Hou1, Peng Zhang2, and Xueru Liang1
1 School of Computer Sci. & Tec., Tianjin University, China
2 School of Computing, The Robert Gordon University, UK
Learning to rank for information retrieval needs some domain experts to label the documents
used in the training step. It is costly to label documents for different research areas. In this
paper, we propose a novel method which can be used as a cross-domain adaptive model based
on importance weighting, a common technique used for correcting the bias or discrepancy. Here
we use “cross-domain” to mean that the input distribution is different in the training and testing
phases. Firstly, we use Kullback-Leibler Importance Estimation Procedure (KLIEP), a typical
method in importance weighing, to do importance estimation. Then we modify AdaRank so that
it becomes a transductive model. Experiments on OHSUMED show that our method performs
better than some other state-of-the-art methods. 13:55– 14:10 Afternoon Aug. 12 Room A Paper No: 1579
Succinct Initialization Methods for Clustering Algorithms
Xueru Liang1, Shangkun Ren1, and Lei Yang2
1 School of Computer Sci. & Tec., Tianjin University, China
2 College of Science, China University of Petroleum, China
In this paper, we focus on the problem of unsupervised clustering of a data-set. We introduce
the traditional K-Means (K-means) cluster analysis and fuzzy C-means (FCM) cluster analysis
of the principles and algorithms process at first, then a novel method to initialize the cluster
centers is proposed. The idea is that the cluster centers’ distribution should be as evenly as
possible within the input field. A “Two-step method” is used in our evolutionary models, with
evolutionary algorithms to get the initialized centers, and traditional methods to get the final
results. Experiment results show our initialization method can speed up the convergence, and in
some cases, make the algorithm performs better. 14:10 – 14:25 Afternoon Aug. 12
Balanced-sampling-based Heterogeneous SVR Ensemble for Business Demand
Forecasting
Yue Liu1, Wang Wei1,Kang Wang1, Zhenjiang Liao1 and Jun-jun Gao2
1 School of Computer Engineering & Science, Shanghai University,
22
Room A Paper No: 1901
Shanghai, 200072, China
2 Sydney Institute of Language and Commerce, Shanghai University,
Shanghai, 201800, China
An accurate demand forecasting model has academic and practical significance to supply chain
management. However, multi-source data and error data have great effect on the demand
prediction accuracy. Therefore, a balanced-sampling-based ensemble of heterogeneous support
vector regression forecasting method named BS-EnHSVR (Balanced-Sampling-based
Ensemble of Heterogeneous SVR) is proposed in this paper to improve the prediction accuracy
by employing balanced sampling and heterogeneous ensemble learning techniques. Training
dataset is firstly classified to different clusters by using clustering algorithm, and then sample
data from each cluster equally to generate training subset for training different individual SVR
models with different training parameters for ensemble. Experimental results on beer sales
show that the proposed method has good usability and generalization ability. 14:25– 14:40 Afternoon Aug. 12 Room A Paper No: 1637
Adaptive Weighted Fusion of Local Kernel Classifiers for Effective Pattern Classification
Shixin Yang1, Wangmeng Zuo1, Lei Liu1, Yanlai Li1, David Zhang1,2
1 Biocomputing Research Centre, School of Computer Science and Technology,
Harbin Institute of Technology, Harbin 150001, China
2 Department of Computing, The Hong Kong Polytechnic University, Hung Hom,
Kowloon, Hong Kong
The theoretical and practical virtual of local learning algorithms had been verified by the
machine learning community. The selection of the proper local classifier, however, remains a
challenging problem. Rather than selecting one single local classifier, in this paper, we propose
to choose several local classifiers and use adaptive fusion strategy to alleviate the choice
problem of the proper local classifier. Based on the fast and scalable local kernel support vector
machine (FaLK-SVM), we adopt the self-adaptive weighting fusion method for combining
local support vector machine classifiers (FaLK-SVMa), and provide two fusion methods,
distance-based weighting (FaLK-SVMad) and rank-based weighting methods (FaLK-SVMar).
Experimental results on fourteen UCI datasets and three large scale datasets show that
FaLK-SVMa can chieve higher classification accuracy than FaLK-SVM. 14:40 – 14:55 Afternoon Aug. 12 Room A Paper No: 1746
Asymmetric Constraint Optimization Based Adaptive Boosting for Cascade Face Detector
Jia-Bao Wen1, 2 and Yue-Shan Xiong1
1 College of Computer, National University of Defense Technology, Changsha, Hunan,
China
2 College of Information Science and Engineering, Hunan University, Changsha,
Hunan, China
A novel variant of AdaBoost named AcoBoost is proposed to directly solve the asymmetric
constraint optimization problem for cascade face detector using a two-stage feature selection
approach. In the first stage, many candidate features are picked out by minimizing the weighted
error. In the second stage, the optimal feature is singled out by minimizing the asymmetric
constraint error. By doing so, the convergence rate is greatly speeded up. Besides, a new sample
set called selection set is added into AcoBoost to prevent overfitting on the training set, which
23
ensures good enough generalization ability for AcoBoost. The experimental results on building
several upright frontal cascade face detectors show that the AcoBoost based classifiers have
much better convergence ability and slightly worse generalization ability than the AdaBoost
based ones. Some AcoBoost based cascade face detectors have satisfactory performance on the
CMU+MIT upright frontal face test set. 14:55 – 15:10 Afternoon Aug. 12 Room A Paper No: 1768
A Linear Regression Model for Nonlinear Fuzzy Data
Juan C. Figueroa-Garcia1 and Jesus Rodriguez-Lopez2
1 Universidad Distrital Francisco Jos´e de Caldas, Bogot´a - Colombia.
2 Universidad Distrital Francisco Jos´e de Caldas, Bogot´a - Colombia.
Fuzzy linear regression is an interesting tool for handling uncertain data samples as an
alternative to a probabilistic approach. This paper sets forth uses a linear regression model for
fuzzy variables; the model is optimized through convex methods. A fuzzy linear programming
model has been designed to solve the problem with nonlinear fuzzy data by combining the
fuzzy arithmetic theory with convex optimization methods. Two examples are solved through
different approaches followed by a goodness of fit statistical analysis based on the measurement
of the residuals of the model. 15:10 – 15:25 Afternoon Aug. 12 Room A Paper No: 1856
Translation Model of Myanmar Phrases for Statistical Machine Translation
Thet Thet Zin1, Khin Mar Soe2 and Ni Lar Thein3
1,3University of Computer Studies, Yangon, Myanmar2Natural Language Processing
Laboratory
University of Computer Studies, Yangon, Myanmar
In this paper, we present a translation model which uses syntactic structure and morphology of
Myanmar language to improve Myanmar to English machine translation system. This system is
implemented as a subsystem of Myanmar to English translation system and based on statistical
approach by using Myanmar-English Bilingual corpus. It also uses two types of information:
language model and translation model. The source language model is based on N-gram method
to extract phrases from segmented Myanmar sentences and the translation model is based on
syntactic structure, morphology of Myanmar language and Bayes rule to reformulate the
translation probability. Experimental results showed that the proposed system gets a
BLEU-score improvement of more than 22.08% in comparison with baseline SMT system. 15:25 – 15:40 Afternoon Aug. 12 Room A Paper No: 1679
A Multi-Agent Reinforcement Learning with Weighted Experience Sharing
Yu Lasheng and Abdulai Issahaku
School of Information Science and Engineering,Central South University,China
Reinforcement Learning, also sometimes called learning by rewards and punishments is the
problem faced by an agent that must learn behavior through trial-and-error interactions with a
dynamic environment [1]. With repeated trials however, it is expected that the agent learns to
perfect its behavior overtime. In this paper we simulate the reinforcement learning process of a
mobile agent on a grid space and examine the situation in which multiple reinforcement
learning agents can be used to speed up the learning process by sharing their Q-values. We
propose a sharing method which takes into consideration the weight of the experience acquired
by each agent on the occasion of visiting a state and taking an action.
24
15:40-15:50
Room A Coffee Break
Machine Learning Theory and Methods II Chair Rudy Rotili, Vincenzo Di Lecce 15:50 – 16:05 Afternoon Aug. 12 Room A Paper No: 1484
Stability Analysis of Neutral Systems with Distributed Delays
Duyu Liu, Qinzhen Huang
College of Electrical and Information Engineering, Southwest University for
Nationalities.
The stability of linear neutral systems with distributed de- lays is studied in this paper. A new
neutral and discrete delays decompos-tion Lyapunov functional method is proposed. Some
stability conditions are derived. A numerical example illustrates that the stability criteria in this
paper is less conservative than the existing ones. In addition, by using argumented Lyapunov
functional, this method allows the coefficient matrix of the neutral term to have time-varying
uncertainties. 16:05 – 16:20 Afternoon Aug. 12 Room A Paper No: 1510
Interval Type-2 Fuzzy Markov Chains: Type Reduction
Juan C. Figueroa-Garc´ıa1, Dusko Kalenatic2, and Cesar Amilcar Lopez3
1 Universidad Distrital Francisco Jos´e de Caldas, Bogot´a - Colombia.
2 Universidad de La Sabana, Ch´ıa - Colombia
3 Universidad Distrital Francisco Jos´e de Caldas, Bogot´a - Colombia
This paper shows an application of Type-reduction algorithms for computing the steady state of
an Interval Type-2 Fuzzy Markov Chain (IT2FM). The IT2FMapproach is an extension of the
scope of a Type-1 fuzzy markov chain (T1FM) that allows to embed several Type-1 fuzzy sets
(T1FS) inside its Footprint of Uncertainty. In this way, a finite state Fuzzy Markov Chain
process is defined on an Interval Type-2 Fuzzy environment, finding their limiting properties
and its Type-reduced behavior. To do so, two examples are provided. 16:20 – 16:35 Afternoon Aug. 12 Room A Paper No: 1172
Structural Fault Diagnosis of Rotating Machinery Based on Distinctive Frequency
Components and Support Vector Machines
Hongtao Xue1, Huaqing Wang2, Liuyang Song2 and Peng Chen1
1 Graduate School of Bioresources, Mie University
2 Beijing University of Chemical Technology
In the field of rotating machinery diagnosis using traditional intelligent diagnosis method, the
state judgment and fault detection are usually carried out by symptom parameters (SPs).
However, it is difficult to find the general and highly sensitive SPs for rotating machinery
diagnosis. Intelligent methods, such as neural networks, genetic algorithms, etc., often cannot
converge when being trained. In order to solve these problems, this paper proposes a new
intelligent diagnosis based on distinctive frequency components (DFCs) and support vector
machines (SVMs) which can be used to detect faults and recognize fault types of rotating
machinery. The method has been applied to detect the structural faults of rotating machinery,
25
and the efficiency of the method is verified by practical examples. 16:35 – 16:50 Afternoon Aug. 12 Room A Paper No: 1036
Probe into Principle of Expert System in Psychological Warfare
Li Shouqi, Long Fangcheng and Wang Yongchang
National University of Defense Technology, China
This paper studies principles, characteristics and process of psychological warfare and the
essentials of artificial intelligence. By providing the theoretical frame of expert system in
psychological warfare (ESPW) based on production rule, this paper makes breakthroughs on the
combination of artificial intelligence and psychological warfare. This theoretical frame is the
foundation of ESPW and it covers production rule set, fact database, knowledge reason tree,
reason machine principle and other contents. 16:50 – 17:05 Afternoon Aug. 12 Room A Paper No: 1653
Real-Time Speech Recognition in a Multi-Talker Reverberated Acoustic Scenario
Rudy Rotili1, Emanuele Principi1, Stefano Squartini1, and Bjorn Schuller2**
1 A3LAB, Department of Biomedics, Electronics and Telecommunications,Universit a
Politecnica delle Marche, Via Brecce Bianche 1, 60131 Ancona, Italy
2 Institute for Human- � �Machine Communication Technische Universit at M unchen Arcisstr.
21, 80333, Munich, Germany
This paper proposes a real-time algorithmic framework for Automatic Speech Recognition
(ASR) in presence of multiple sources in reverberated environment. The addressed real-life
acoustic scenario denitely asks for a robust signal processing solution to reduce the impact of
source mixing and reverberation on ASR performances. Here the authors show how the
implemented approach allows to improve recognition accuracies under real-time processing
constraints and overlapping distanttalking speakers. A suitable database has been generated on
purpose, by adapting an existing large vocabulary continuous speech recognition (LVCSR)
corpus to deal with the acoustic conditions under study. 17:05 – 17:20 Afternoon Aug. 12 Room A Paper No: 1436
Syntactic Pattern Recognition from Observations: a Hybrid Technique
Vincenzo Di Lecce1 and Marco Calabrese1*
1Polytechnic of Bari, DIASS Taranto, Italy
This paper presents a novel technique for automated learning from observations. The technique
arranges in a row four traditional pattern recognition approaches (numeric, logic, statistical and
finally syntactic) within a unifying framework. Each processing step is conceived as a
transformation of the input dataset from one state to another. The proposed technique considers
measurable observations as inputs and produces a set of formal rules, i.e., a grammar, as final
output. To this end, a four-state grammar induction process is described in detail by means of a
step-by-step example. As a proof-of-concept for the feasibility of the proposal, references to
early experimental validations are given. Finally, possible comparison with other well-known
approaches is discussed. 17:20 – 17:35 Afternoon Aug. 12 Room A
An Efficient Ensemble Method for Classifying Skewed Data Streams
Juan ZHANG, Xuegang HU, Yuhong ZHANG and Peipei LI,
School of Computer and Information, Hefei University of Technology, Hefei 230009,
China
26
Paper No: 1442
Class distributions of data streams in real application are usually unbalanced, they are hence
called Skewed Data Streams (abbreviated as SDS). However, in the classification of SDS, it is a
challenge for traditional methods because of the difficulty in the recognition of minority classes.
Therefore, many approaches have been proposed to improve the recognition rate of minority
classes, while they are time-consuming. Motivated by this, we propose an efficient Ensemble
method for Classifying SDS called ECSDS. Our algorithm creates multiple classifiers based on
C4.5, and adopts the threshold of F1-value to limit the updating frequency of classifiers.
Meanwhile, it adds misclassified positive instances into the training data to guarantee the
effectiveness of classifiers when updating. Experimental studies demonstrate that our proposed
method enables reducing the time overhead and maintains a good performance on the
classification accuracy. 17:35 – 17:50 Afternoon Aug. 12 Room A Paper No: 1473
An Association Rules Algorithm Based on Kendall-τ
Anping Zeng1,Yongping Huang2
1School of Computer and Information Engineering, Yibin University,Yibin, Sichuan
644007, China
2Computational Physics Key Laboratory of Sichuan Province, Yibin University,Yibin,
Sichuan 644007, China
The disadvantages of Apriori algorithm are firstly discussed. Then, a new measure of kendall-τ
is proposed and treated as an interest threshold. Furthermore, an improved Apriori algorithm
called K-apriori is proposed based on kendall-τ correlation coefficient. It not only can
accurately find the relations between different products in transaction databases and reduce the
useless rules but also can generate synchronous positive rules, contrary positive rules and
negative rules. Experiment has been carried out to verify the effectiveness of the algorithm. The
result shows that the algorithm is effective at discovering the association rules in a sales
management system.
Room B Neural Networks
Chair Lingzhi Wang 13:40 – 13:57 Afternoon Aug. 12 Room B Paper No: 1375
A Novel Nonlinear Neural Network Ensemble Model Using K-PLSR for Rainfall
Forecasting
Chun Meng and Jiansheng Wu
Department of Mathematics and Computer, Liuzhou Teacher College, Liuzhou,
Guangxi, 545004, China
In this paper, a novel hybrid Radial Basis Function Neural Network (RBF–NN) ensemble
model is proposed for rainfall forecasting based on Kernel Partial Least Squares Regression
(K–PLSR). In the process of ensemble modeling, the first stage the initial data set is divided
into different training sets by used Bagging and Boosting technology. In the second stage, these
training sets are input to the RBF–NN models of different kernel function, and then various
27
single RBF–NN predictors are produced. Finally, K–PLSR is used for ensemble of the
prediction purpose. Our findings reveal that the K–PLSR ensemble model can be used as an
alternative forecasting tool for a Meteorological application in achieving greater forecasting
accuracy. 13:57 – 14:14 Afternoon Aug. 12 Room B Paper No: 1463
Oscillatory behavior for a class of recurrent neural networks with time-varying input and
delays
Chunhua Feng1 and Zhenkun Huang2
1College of Mathematical Science, Guangxi Normal University, Guilin, Guangxi,
541004, P. R. China
2Department of Mathematics, Jimei University, Xiamen, 350211, P. R. China
In this paper, the existence of oscillations for a recurrent neural network with time delays
between neural interconnections is investigated. Several simple and practical criteria to
determine the oscillatory behavior are obtained. 14:14 – 14:31 Afternoon Aug. 12 Room B Paper No: 1576
A Saturation Binary Neural Network for Bipartite Subgraph Problem
Cui Zhang1, Li-Qing Zhao2 and Rong-Long Wang2
1Department of Autocontrol, Liaoning Institute of Science and Technology,Benxi, China
2 Graduate School of Engineering, University of Fukui, Bunkyo 3-9-1, Fukui-shi, Japan
In this paper, we propose a saturation binary neuron model and use it to construct a
Hopfield-type neural network called saturation binary neural network to solve the bipartite
sub-graph problem. A large number of instances have been simulated to verify the proposed
algorithm, with the simulation result showing that our algorithm finds the solution quality is
superior to the compared algorithms 14:31 – 14:48 Afternoon Aug. 12 Room B Paper No: 1416
Neural Network Ensemble Model Using PPR and LS-SVR for Stock Market Forecasting
Lingzhi Wang and Jiansheng Wu
Department of Mathematical and Computer Science, Liuzhou Teachers College
Liuzhou, Guangxi, 545004 China
In this study, a novel Neural Network (NN) ensemble model using Projection Pursuit
Regression (PPR) and Least Squares Support Vector Regression (LS–SVR) is developed for
financial forecasting. In the process of ensemble modeling, the first stage some important
economic factors are selected by the PPR technology as input feature for NN. In the second
stage, the initial data set is divided into different training sets by used Bagging and Boosting
technology. In the third stage, these training sets are input to the different individual NN
models, and then various single NN predictors are produced based on diversity principle. In the
fourth stage, the Partial Least Square (PLS) technology is used to choosing the appropriate
number of neural network ensemble members. In the final stage, LS–SVR is used for ensemble
of the NN to prediction purpose. For testing purposes, this study compare the new ensemble
model’s performance with some existing neural network ensemble approaches in terms of the
Shanghai Stock Exchange index. Experimental results reveal that the predictions using the
proposed approach are consistently better than those obtained using the other methods
presented in this study in terms of the same measurements.
28
14:48 – 15:05 Afternoon Aug. 12 Room B Paper No: 1437
On-line Extreme Learning Machine for Training Time-Varying Neural Networks
Yibin Ye, Stefano Squartini, and Francesco Piazza
A3LAB, Department of Biomedics, Electronics and Telecommunications, Universit`a
Politecnica delle Marche, Via Brecce Bianche 1, 60131 Ancona, Italy
Time-Varying Neural Networks(TV-NN) represent a powerful tool for nonstationary systems
identification tasks, as shown in some recent works of the authors. Extreme Learning Machine
approach can train TV-NNs efficiently: the reference algorithm is named ELM-TV and is of
batch-learning type. In this paper, we generalize an online sequential version of ELM to TV-NN
and evaluate its performances in two nonstationary systems identification tasks. The results
show that our proposed algorithm produces comparable generalization performances to
ELM-TV with certain benefits to those applications with sequential arrival or large number of
training data. 15:05 – 15:22 Afternoon Aug. 12 Room B Paper No: 1183
A Partially Connected Neural Evolutionary network for Stock Price Index Forecasting
Didi Wang2,;Pei-Chann Chang1, Jheng-Long Wu1,;Changle Zhou2 1Department of Information Management, Yuan Ze University, Taoyuan 32026 2Cognitive Science Department, Fujian Key Laboratory of the Brain-like Intelligent
Systems, Xiamen University, Xiamen, China
This paper proposes a novel partially connected neural evolutionary model (Parcone)
architecture to simulate the relationship of stock and technical indicators to predict the stock
price index. Different from artificial neural networks, the architecture has corrected three
drawbacks: (1) connection between neurons of is random; (2) there can be more than one
hidden layer; (3) evolutionary algorithm is employed to improve the learning algorithm and
train weights. The more hidden knowledge stored within the historic time series data are needed
in order to improve expressive ability of network. The genetically evolved weights mitigate the
well-known limitations of gradient descent algorithm. In addition, the activation function is not
defined by sigmoid function but sin(x). The experimental results show that Parcone can make
the progress concerning the stock price index and it’s very promising to calculate the predictive
percentage by simulation results of proposed evolutionary system. 15:22 – 15:40 Afternoon Aug. 12 Room B Paper No: 1260
Finite Precision Extended Alternating Projection Neural Network (FPEAP)
Yanfei Wang, Jingen Wang
New Star Research Institute of Applied Technology, 451 Huangshan Road, Hefei,
Anhui, China
The paper studies finite precision Extended Alternating Projection Neural Network (FPEAP)
and its related problems. An improved training method of FPEAP has been present after
considering the finite precision influence on the training method of EAP. Then the mathematical
relation among the factors influencing the association times has been studied. Finally simulation
experiments have been designed and simulation results demonstrate validity of theoretical
analyses.
29
15:40 – 15:50 Room B Coffee Break
Intelligent Control and Automation Chair Hongyan Yan 15:50 – 16:03 Afternoon Aug. 12 Room B Paper No: 1219
H∞output tracking control for neutral delay system with nonlinear perturbations
Cuihong Wang and Huijuan Cao
Department of Mathematics and Computer Science, Shanxi Normal University, Shanxi
041004 China
This paper studies the problem of H∞ output tracking control for neutral delay system with
nonlinear perturbations. Based on a novel augmented Lyapunov functional, a sufficient
condition for the existence of state-feedback controller is presented in terms of linear matrix
inequality, which not only guarantees the stability of closed-loop system, but also ensure the
output of the neutral system tracks the output of a given reference model in the H∞ sense. 16:03 – 16:17 Afternoon Aug. 12 Room B Paper No: 1869
Continuous Finite-Time Observer-Controller Based Speed Regulation of Permanent
Magnet Synchronous Motors
Yan Yan1, Shuanghe Yu1, Zhenqiang Yang2 and Jialu Du1
1School of Information Science and Technology, Dalian Maritime University 116026
Dalian, China
2School of Electrical Engineering, Dalian University of Technology 116024 Dalian,
China
A finite-time speed tracking control for a permanent-magnet synchronous motor (PMSM) using
an unknown load torque disturbance estimation technique is presented. The proposed control
scheme depends on finite-time convergence of a state observer, which estimates motor speed,
acceleration and the unknown load torque. A continuous output feedback controller that can
achieve global finite-time stability for a PMSM is constructed based on a “finite-time separation
principle”. The simulation results show the efficiency of the method. 16:17 – 16:30 Afternoon Aug. 12 Room B Paper No: 1918
Observer-Based Exponential Stability Analysis for Networked Control Systems with
Packet Dropout
Xue Li1, Jia-min Weng2, Dajun Du1 and Haoliang Bai1
1 Department of Automation, School of Mechatronical Engineering and Automation,
Shanghai University, Shanghai, 200072, China
2 Department of Electrical Information Engineering, Henan Institute of Engineering,
Zhengzhou, 451191, China
This paper is concerned with observer-based exponential stability for networked control
systems (NCSs) with data packet dropout. Firstly, the data packet dropouts both the
sensor-to-controller and the controller-to-actuator are considered, which are modeled as two
independent Bernoulli distributed white sequences respectively. Then an observer is designed to
estimate the system state, and an augmented model for NCSs is proposed. Finally, based on
30
Lyapunov stability theory combined with linear matrix inequalities (LMIs) techniques, a su
cient condition is derived for NCSs to be exponentially mean-square stable. 16:30 – 16:44 Afternoon Aug. 12 Room B Paper No: 1449
Three Levels Intelligent Incident Detection Algorithm of Smart Traffic in the Digital City
Hongyan Yan1, Xiaojuan Zhang2, Hongxia Xu3
1 School of Computer and Communication Engineering, Zhengzhou University of Light
Industry, Henan, China
2 Department of Foreign Language, Zhengzhou Tourism College, Henan, China
3 Art Design Department, Jiyuan Vocational and Technical College, Jiyuan, China
In the paper, digital city is studied first, and then the smart traffic system is proposed. After that,
a high-efficiency three levels intelligent incident detection algorithm is designed in detail. A
better efficiency analysis by detecting rate, false alarm rate, and average detecting time, is
obtained by simulation and experiment on Zhengzhou Ring Highway and BRT system. This
algorithm not only fits flat plain, but also ring roads, therefore it can help city planning
administrator implement the maximizing of traffic flows. 16:44 – 16:57 Afternoon Aug. 12 Room B Paper No: 1749
Improvement of Path Planning in Mobile Beacon Assisted Positioning
Ji-Rui Li1 and Kai Yang2
1 Information Engineer Department,Henan Vocational and Technical Institute Lecturer,
Zhengzhou , China
2Computer Department,Armed Police Command College in Zhengzhou Lecturer,
Zhengzhou , China
In static Wireless sensor networks position, at present using a mobile beacon assisted
positioning mechanism. However, in assisted positioning of the mobile beacon, the mobile path
of beacon nodes has a great influence to positioning performance, the existing mobile assisted
positioning algorithm consider less to path planning, for its lack, paper proposes a more flexible
dynamic heuristic path planning method, with more practical for the application of irregular
network topology. 16:57– 17:10 Afternoon Aug. 12 Room B Paper No: 1939
Sliding Mode Observer Based Anti-Windup PI Speed Controller for Permanent Magnet
Synchronous Motors
Shuanghe Yu1 , Zhenqiang Yang2, Jialu Du1 and Jingcong Ma1
1 School of Information Science and Technology, Dalian Maritime University 116026
Dalian, China
2 School of Electrical Engineering, Dalian University of Technology 116024 Dalian,
China
This paper proposes a new sliding mode observer (SMO) based anti-windup PI speed control
strategy for Permanent Magnet Synchronous Motor (PMSM). A SMO is constructed to estimate
the back-electromotive-force (EMF) signals robustly, and then the position and speed of a
PMSM can be calculated according to the back-EMF equations. Based on the estimated
position and speed of PMSM, four different anti-windup PI strategies are used and compared
for speed control of PMSM, The simulation results show the superior control performance
compared with conventional schemes. 17:10 – 17:23
Robotic Wheelchair Moving with Caregiver Collaboratively
31
Afternoon Aug. 12 Room B Paper No: 1309
Yoshinori Kobayashi1, Yuki Kinpara1, Erii Takano1, Yoshinori Kuno1,Keiichi Yamazaki1,
and Akiko Yamazaki2
1Department of Information and Computer Sciences, Saitama University,255
Shimo-Okubo, Sakura-ku, Saitama city, Saitama 338-8570, Japan. 2School of Media
Science, Tokyo University of Technology,1404-1 Katakuramachi, Hachioji city, Tokyo
192-0982, Japan
This paper introduces a robotic wheelchair that can automatically move alongside a caregiver.
Because wheelchair users are often accompanied by caregivers, it is vital to consider how to
reduce a caregiver's load and support their activities, while simultaneously facilitating
communication between the caregiver and the wheelchair user. Moreover, a sociologist pointed
out that when a wheelchair user is accompanied by a companion, the latter is inevitably seen by
others as a caregiver rather than a friend. In other words, the equality of the relationship is
publicly undermined when the wheelchair is pushed by a companion. Hence, we propose a
robotic wheelchair able to move alongside a companion, and facilitate easy communication
between the companion and the wheelchair user. Laser range sensors are used for tracking the
caregiver and observing the environment around the wheelchair. To confirm the effectiveness of
the wheelchair in real-world situations, we conducted experiments at an elderly care center in
Japan. Results and analyses are also reported in this paper. 17:23 – 17:36 Afternoon Aug. 12 Room B Paper No: 1415
Exploration Strategy Related Design Considerations of WSN-Aided Mobile Robot
Exploration Teams
Gurkan Tuna1, Kayhan Gulez2, Vehbi Cagri Gungor3 and Tarik Veli Mumcu2
1Trakya University, Vocational College of Technical Sciences, Department of Computer
Programming, Edirne, [email protected]
2Yildiz Technical University, Electrical-Electronics Eng. Faculty, Control and
Automation Eng. Dept., Istanbul, TURKEY
3Bahcesehir University, Faculty of Engineering, Department of Computer Engineering,
Istanbul, TURKEY [email protected]
This paper presents a novel approach to mobile robot exploration. In this approach, mobile
robots send their local maps to the central controller and coordinate with each other using a
wireless sensor network (WSN). Different from existing rendezvous point-based exploration
strategies, the use of a WSN as the communication media allows quick and cost-effective
exploration and mapping of an unknown environment. Overall, this paper introduces
WSN-aided mobile robot exploration strategy and shows comparative performance evaluations
using the Player/Stage simulation platform. Here, our main goal is to present potential
advantages of WSN-aided mobile robot exploration for Simultaneous Localization and
Mapping (SLAM). 17:36 – 17:50 Afternoon Aug.12 Room B Paper No: 1196
Robust Controller Design for Main Steam Pressure Based on SPEA2
Shuan Wang1 , Dapeng Hua1 , Zhiguo Zhang1, Ming Li1 , Ke Yao1, Zhanyou Wen 2
Henan Nanyang Power Supply Company, Nanyang, Henan, 473000, China
Main steam pressure is an important physical quantity that reflects the energy supply-demand
relationship between the boiler and turbine. It has a significant role in the unit operation.
32
Because boiler burning behavior varies greatly and the model of main steam pressure is of
highly uncertainty, conventional control method can not obtain the expected control effect. In
order to improve system control quality, a robust controller for main steam pressure is designed
by using the H∞ mixed sensitivity approach in this paper. To better meet the site requirements,
adding more restrictions in design, author innovation to put such a complex issue into
multiobjective optimization problems. SPEA2 (The Strength Pareto Evolutionary Algorithm 2)
are used to optimize the parameters of weighing functions in order to search for the H∞
controller which meets the time-and frequency-domain indexes. Simulation results show that
the design of the main steam pressure control system has an excellent robust stability and
dynamic quality.
Room C Intelligent Computing in Bioinformatics & Bio-medical
Engineering Chair M. Michael Gromiha & Jingyan Wang 13:40 – 13:55 Afternoon Aug. 12 Room C Paper No: 1987
Evolving Gene Regulatory Networks: A Sensitivity-Based Approach
Yu-Ting Hsiao Wei-Po Lee
Department of Information Management National Sun Yat-sen University Kaohsiung,
Taiwan
Constructing genetic regulatory networks (GRNs) from expression data is one of the most
important issues in systems biology research. To automate the procedure of network
construction, we develop an evolution framework to infer the S-system network models. Our
framework mainly includes a sensitivity analysis method and a hybrid GA-PSO method to infer
appropriate network parameters. To validate the proposed methods, experiments have been
conducted and the results show the promise of our approach. 13:55– 14:10 Afternoon Aug. 12 Room C Paper No: 1690
A Faster Haplotyping Algorithm based on Block Partition and Greedy Ligation Strategy
Xiaohui Yao1,2, Yun Xu1,2,�, Jiaoyun Yang1,2, and Wenhua Cheng1,2
1 Department of Computer Science, University of Science and Technology of
China,Hefei, Anhui 230026, China
2 Anhui Province-MOST Co-Key Laboratory of High Performance Computing and Its
Application, Hefei, Anhui 230027, China
Haplotype played a very important role in the study of some disease gene and drug response
tests over the past years. However, it is both time consuming and very costly to obtain
haplotypes by experimental way. Therefore haplotype inference was proposed which deduce
haplotypes from the genotypes through computing methods. Some genetic models were
presented to solve the haplotype inference problem, and Maximum Parsimony model was one
of them, but at present the methods based on this principle are either simple greedy heuristic or
exact ones, which are adequate only for moderate size instances. In this paper, we presented a
faster greedy algorithm named FHBPGL applying partition and ligation strategy. Theoretical
33
analysis shows that this strategy can reduce the running time for large scale dataset and
following experiments demonstrated that our algorithm gained comparable accuracy compared
to exact haplotyping algorithms with less time. 14:10– 14:25 Afternoon Aug. 12 Room C Paper No: 1621
Predicting long-term vaccine e cacy against metastases using agents
Marzio Pennisi1, Dario Motta1, Alessandro Cincotti 2, and Francesco Pappalardo1
University of Catania, Catania, Italy;School of Information Science, Japan Advanced
Institute of Science and Technology, Japan
To move faster from preclinical studies (experiments in mice) towards clinical phase I trials
(experiments in advanced cancer patients), the chance to predict the outcome of longer
experiments represents a key step.We use the MetastaSim model to predict the long-term effects
of the Triplex vaccine against metastases. To this end we simulate follow-ups of two and three
of three months (equivalent approximately to 5.83 and 8.75 years in humans) to compare the
long-term e cacy of the best protocol used “in vivo" against the one found by the MetastaSim
model. We also check the efficacy of these two protocols by delaying the time of the rst
administration, in order to catch up the maximum time delay between the appearing of
metastases and the administration of the vaccine needed to guarantee reasonable treatment
efficacy. 14:25 – 14:40 Afternoon Aug. 12 Room C Paper No: 1984
Using 2D Principal Component Analysis to Reduce Dimensionality of Gene Expression
Profiles for Tumor Classification
Shu-Lin Wang1, Min Li1 and Hongqiang Wang2
1 College of Information Science and Engineering, Hunan University, Changsha,
Hunan, 410082, China
2 The Intelligent Computing Laboratory, Hefei Institute of Intelligent Machines, Chinese
Academy of Sciences, Hefei, Anhui, 230031, China
In the last ten years, numerous methods have been proposed for accurate classification of tumor
subtype based on gene expression profiles (GEP). Among these methods, feature extraction
methods play an important role in constructing classification model. However, traditional
methods view a gene expression sample as 1D vector, which does not sufficiently utilize the
correlation and structure information among many genes. We, therefore, introduce 2D principal
component analysis (2DPCA) to extract features for tumor classification by converting 1D
sample vector into 2D sample matrix. To evaluate its performance, we perform a series of
experiments on four tumor datasets. The experimental results indicate that the obtained
performance by using 2DPCA is superior to the classic principal component analysis. 14:40 – 14:55 Afternoon Aug. 12 Room C Paper No: 1494
Structure-Function relationship in Olfactory Receptors
M. Michael Gromiha1,*, R. Sowdhamini2 and K. Fukui3
1 Department of Biotechnology, Indian Institute of Technology Madras, Chennai 600
036, Tamilnadu, India,
2 National Center for Biological Sciences, Bangalore, India
3 Computational Biology Research Center, National Institute of Advanced Industrial
Science and Technology, 2-4-7 Aomi, Koto-ku, Tokyo 135-0064, Japan
Olfactory receptors are key components in signal transduction. The sequence and structural
34
analysis of olfactory receptors provides deep insights to understand their function. In this work,
we have systematically analyzed the relationship between various physical, chemical, energetic
and conformational properties of amino acid residues, and the change of half maximal effective
concentration (EC50) due to amino acid substitutions. We observed that the odorant molecule
(lignad) as well as amino acid properties are important for EC50. The inclusion of
neighboring/surrounding residues information of the mutants enhanced the correlation. Further,
amino acid properties have been combined systematically and we obtained a correlation of
0.90-0.98 with functional data for different (goldfish, mouse and human) olfactory receptors. 14:55 – 15:10 Afternoon Aug. 12 Room C Paper No: 1294
Inferring protein-protein interactions based on sequences and interologs in
Mycobacterium tuberculosis
Zhi-Ping Liu1, Jiguang Wang2, Yu-Qing Qiu3, Ross K.K. Leung4, Xiang-Sun Zhang3,
Stephen K.W. Tsui4 and Luonan Chen1
1 Key Laboratory of Systems Biology, Shanghai Institutes for Biological Sciences,
Chinese Academy of Sciences, Shanghai 200031, China
2 Beijing Institute of Genomics, Chinese Academy of Sciences, Beijing 100029, China
3 Academy of Mathematics and Systems Science, Chinese Academy of Sciences,
Beijing 100190, China
4 Hong Kong Bioinformatics Centre, The Chinese University of Hong Kong, Shatin N.
T., Hong Kong, China
Mycobacterium tuberculosis is a pathogenic bacterium that poses serious threat to human
health. Inference of the protein interactions of M. tuberculosis will provide cues to understand
the biological processes in this pathogen. In this paper, we constructed an integrated M.
tuberculosis H37Rv protein interaction network by machine learning and ortholog-based
methods. Firstly, we developed a support vector machine (SVM) method to infer the protein
interactions by gene sequence information. We tested our predictors in Escherichia coli and
mapped the genetic codon features underlying protein interactions to M. tuberculosis.
Moreover, the documented interactions of other 14 species were mapped to the proteome of M.
tuberculosis by the interolog method. The ensemble protein interactions were then validated by
various functional linkages i.e., gene coexpression, evolutionary relationship and functional
similarity, extracted from heterogeneous data sources. 15:10 – 15:25 Afternoon Aug. 12 Room C Paper No: 1243
MiRaGE: Inference of gene expression regulation via microRNA transfection II
Y-h. Taguchi1 and Jun Yasuda2
1 Department of Physics, Chuo University, Tokyo 112-8551, Japan,
2 COE Fellow, Graduate School of Medicine, Tohoku University, Sendai 980-8575,
Japan
How each microRNA regulates gene expression is unknown problem. Especially, which gene is
targeted by each microRNA is mainly depicted via computational method, typically without
biological/experimental validations. In this paper, we propose a new computational method,
MiRaGE, to detect gene expression regulation via miRNAs by the use of expression profile data
and miRNA target prediction. This method is tested to miRNA transfection experiments to
tumor cells and succeeded in inference of transfected miRNA as only one miRNA with
significant P-values for the first time.
35
15:25– 15:40 Afternoon Aug. 12 Room C Paper No: 1983
Semi-supervised Protein Function Prediction via Sequential Linear Neighborhood
Propagation
Jingyan Wang, Yongping Li �, Ying Zhang, and Jianhua He
Shanghai Institute of Applied Physics, Chinese Academy of Sciences, 2019 Jialuo
Road, Jiading District, Shanghai 201800, P. R. China.
Predicting protein function is one of the most challenging problems of the post-genomic era.
The development of experimental methods for genome scale analysis of molecular interaction
networks has provided new approaches to inferring protein function. In this paper we introduce
a new graph-based semi-supervised classi cation algorithm Sequential Linear Neighborhood
Propagation (SLNP), which addresses the problem of the classi cation of partially labeled
protein interaction networks. The proposed SLNP rstly constructs a sequence of node sets
according to their shortest distance to the labeled nodes, and then predicts the function of the
unlabel proteins from the set closer to labeled one, using Linear Neighborhood Propagation. Its
performance is assessed on the Saccharomyces cerevisiae PPI network data sets with good
results compared with three current state-of-the-art algorithms. 15:40 – 15:50
Room C Coffee Break
Intelligent Computing in Pattern Recognition Chair Linlin Shen 15:50 – 16:05 Afternoon Aug. 12 Room C Paper No: 1591
Smart Sensors: A Holonic Perspective
Vincenzo Di Lecce , Marco Calabrese
Politecnico di Bari, DIASS Taranto, Italy
This work introduces a novel perspective in the study of smart sensors technology. The final
aim is to develop a new methodology that supports the conception, design and implementation
of complex sensor-based systems in a more structured and information-oriented way. A smart
sensor can be considered as a hardware/software transducer able to bring the measured physical
signal(s) at an application level. However, when viewed through the lens of artificial
intelligence, sensor ‘smartness’ appears to stay in between merely transduction and complex
post-processing, with the boundary purposely left blurry and undetermined. Thanks to the
recent literature findings on the so-called ‘holonic systems’, a more precise characterization and
modeling of the smart sensor is provided. A ‘holon’ is a bio-inspired conceptual and
computational entity that, as a cell in a living organism, plays the two roles of a part and a
whole at the same time. To bring the right evidence of to the advantages of the holonic
approach, an example smart application and a related prototype implementation for the
disambiguation of low-cost gas sensor responses is shown. The proposed approach unravels the
inherent complexity of the disambiguation problem by means of a scalable architecture entirely
based on holonic-inspired criteria. Furthermore, the overall setup is economically competitive
with other high-selective (hence high-cost) sensor-based solutions. 16:05– 16:20
Texture Classification Based on Contourlet Subband Clustering
36
Afternoon Aug. 12 Room C Paper No: 1541
Yongsheng Dong and Jinwen Ma
Department of Information Science, School of Mathematical Sciences and LMAM,
Peking University, Beijing, 100871, China
In this paper, we propose a novel texture classification method based on feature extraction
through c-means clustering on the contourlet domain. In particular, all the features representing
each contourlet subband are extracted by a c-means clustering standard algorithm. By
investigating these features, we use the weighted L1-norm for comparing the features of the two
corresponding subbands of two images and de¯ne a new distance between two images.
According to the new distance, a k-Nearest Neighbor (kNN) classifier is utilized to perform
texture classification (TC), and experimental results reveal that our proposed approach
outperforms two current state-of-the-art texture classi¯cation approaches. 16:20– 16:35 Afternoon Aug. 12 Room C Paper No: 1194
Face Recognition from Visible and Near-Infrared Images Using Boosted Directional
Binary Code
Linlin Shen, Jinwen He, Shipei Wu, and Songhao Zheng
School of Computer Science & Software Engineering Shenzhen University, China
Pose and illuminations remain great challenges to current face recognition technique. In this
paper, visible image (VI) and near-infrared image (NIR) are fused for performance
improvement. When directional binary code is adopted as feature representation, AdaBoost
algorithm and the cascade structure are used for classification. Fusion is done at decision level
and classification scores are normalized using three different rules, i.e. Min-Max, Z-Score and
Tanh-Estimators. Experimental results suggest that the proposed algorithm using VI achieve
better performance than NIR when pose and expression variations are present. However, NIR
shows much better robustness against illumination and time difference than VI. Due to the
complementary information available in two image modalities, fusion of NIR and VI further
improves the system performance. 16:35– 16:50 Afternoon Aug. 12 Room C Paper No: 1483
A Systematic Algorithm for Fingerprint Image Quality Assessment
Min Wu1,A Yong2, Tong Zhao2, Tiande Guo2
1 School of information Sciences and Engineering, Graduate University of Chinese
Academy of Sciences, Beijing, China
2 School of mathematical Sciences, Graduate University of Chinese Academy of
Sciences, Beijing, China
The fingerprint image quality is a key factor on the match results since it will cause spurious
and missed minutiae when matching with the low quality images. It is important to estimate the
image quality to guide the feature extraction and matching. In this paper we investigate the
specifications that can reflect the image quality such as orientation coherence, core position and
so on. We define a quasi core as a stable point to examine the validity of the captured position.
We apply the idea of penalty function in the optimization theory to combine the specifications
to get a quality score. The method is robust since it investigates the quality specifications
entirely. The testing results on FVC database are given to verify the feasibility and
effectiveness. 16:50 – 17:05
Adaptive Variance based Sharpness Computation for Low Contrast Images
37
Afternoon Aug. 12 Room C Paper No: 1694
Xin Xu1,2, Yinglin Wang2, Jinshan Tang1, Xiaolong Zhang1 and Xiaoming Liu1
1 School of Computer Science and Technology, Wuhan University of Science and
Technology, Wuhan 430081, China
2 Department of Computer Science and Engineering, School of Electronic, Information
and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
Low contrast images are easily suffering from noise effect. As a result, it can witness many
local false peaks in the graph of sharpness function. However, the presence of many local false
peaks hinders the camera's passive auto-focus system to perform its function in locating the
focused peak. This paper presents an improved variance based sharpness computation which
can adapt to various degrees of noise. The proposed sharpness computation can bring in the
local false peaks generated by noise influence, and therefore produce a well defined focused
peak standing for the best focused image. The experimental results from several image
sequences validate the effectiveness of our proposed method. 17:05 – 17:20 Afternoon Aug. 12 Room C Paper No: 1860
Mass segmentation in mammograms based on improved level set and watershed
algorithm
Jun Liu, Xiaoming Liu, Jianxun Chen* and J Tang*
College of Computer Science and Technology, Wuhan University of Science and
Technology, Wuhan, Hubei, China
In this paper, a new mass segmentation algorithm is proposed. In the new proposed algorithm, a
fully automatic marker-controlled watershed transform is first proposed to segment the mass
region roughly, and then a level set is used to refine the segmentation. The new algorithm
combines the advantages of both methods. The combination of the watershed based
segmentation and level set method can improve the efficiency of the segmentation. Images from
DDSM were used in the experiments and the results show that the new algorithm can improve
the accuracy of mass segmentation. 17:20– 17:35 Afternoon Aug. 12 Room C Paper No: 1657
Face recognition system robust to occlusion
Mohit Sharma, Surya Prakash, and Phalguni Gupta
Department of Computer Science and Engineering, Indian Institute of Technology
Kanpur Kanpur 208016, INDIA
This paper presents an ecient face recognition system which can handle partial occlusion in
both training and test image sets. We hybridize Gabor lter with Eigen faces which make use of
localization e ect of Gabor lter and whole appearance e ect of Eigen faces. It has been tested on
AR database which contains naturally occluded face images. 17:35 – 17:50 Afternoon Aug. 12 Room C Paper No: 1411
Computing Technique in Ancient India
Chitralekha Mehera
Institute of Science Education, The University of Burdwan, Burdwan, West Bengal,
India
This paper deals with the Sutra “Nikhilam Navatascaramam Dasatah” of Vedic Mathematics in
computing mathematical problems viz., 10’s complement, multiplication tables, addition,
subtraction, multiplication and division quickly considering various examples. The problem
38
solving techniques using the Sutra not only minimize the computational time but also seem to
the students as an interesting and effective mathematical learning. The paper also explores the
algebraic explanations of these techniques.
Saturday, Morning, August 13
Room A Evolutionary Learning & Genetic Algorithms & Optimization
Chair Roman Neruda, Syeda Darakhshan Jabeen 8:00 – 8:13 Morning Aug. 13 Room A Paper No: 1563
Local Meta-models for ASM-MOMA
Martin Pil´at1,2 and Roman Neruda2
1 Department of Theoretical Computer Science and Mathematical Logic, Faculty of
Mathematics and Physics, Charles University in Prague, Malostransk´e n´amˇest´ı 25,
Prague, Czech Republic [email protected]
2 Institute of Computer Science, Academy of Sciences of the Czech Republic, Pod
Evolutionary algorithms generally require a large number of objective function evaluations
which can be costly in practice. These evaluations can be replaced by evaluations of a cheaper
meta-model of the objective functions. In this paper we describe a multi-objective memetic
algorithm utilizing local distance based meta-models. This algorithm is evaluated and compared
to standard multi-objective evolutionary algorithms as well as a similar algorithm with a global
meta-model. The number of objective function evaluations is considered, and also the
conditions under which the algorithm actually helps to reduce the time needed to find a solution
are analyzed.
8:13 – 8:26 Morning Aug. 13 Room A Paper No: 1453
A restrained optimal perturbation method for solving the inverse problem in reverse
process of convection diffusion equation
Bo Wang1,2, Guang-an Zou2, and Peng Zhao2
1 Institute of Applied Mathematics, Henan university, Kaifeng, China
2 College of Mathematics and Information Science, Henan university, Kaifeng, China
In this paper, a new approach of the restrained optimal perturbation method is firstly proposed
to study the inverse problem in the reverse process of the one-dimensional convection diffusion
equation, the idea of this method is brand new that in search for the optimal perturbation value
by the given initial estimate, for determining the initial distribution based on the overspecified
data, and the initial estimates plus optimal perturbation value can be treated as the final initial
distribution, in order to overcome the ill-posedness of this problem, a regularization term is
introduced in the objective functional. Numerical examples will be given, and the results show
that our method is effective.
39
8:26– 8:39 Morning Aug. 13 Room A Paper No: 1007
Vibration Control of A Vehicle Using Hybrid Genetic Algorithm
Syeda Darakhshan Jabeen and Rathindra Nath Mukherjee
Department of Mathematics, University of Burdwan,
Burdwan-713104, India.
In this paper a new hybrid method has been proposed for solving the suspension design
problem. A two-dimensional model of a car with linear passive suspension system and with two
passengers has been considered. The vibration, experienced by the passengers due to road bump
during vehicle motion has been minimized in time domain, by applying the proposed method.
Moreover, the suspension parameters have been determined which satisfy performance as per
ISO standards. The solutions/ parametric values so obtained have been further compared with
the existing suspension parameters. 8:39 – 8:52 Morning Aug. 13 Room A Paper No: 1079
Dynamics of a Two Prey One Predator Fishery with Low Predator Density
T. Das1, R. N. Mukherjee2, K.S. Chaudhuri3
1 Department of Mathematics, University Institute of Technology, Burdwan University,
Burdwan –713104, India
2 Retd.Professor, Department of Mathematics, Burdwan University, Burdwan –
713104,India
3 Professor, Department of Mathematics, Jadavpur University, Kolkata –700032
The present model is concerned with a multispecies fishery with two prey species both of which
obeys the logistic law of growth and one predator species whose density is low. The predator
consumes one of the prey species more intensively than the other because of its availability. For
the predator species the growth function is taken as the model described by Smith. We assume
that both the prey species are subjected to harvesting while the predator species is excluded
from harvesting due to its low density. In this model we consider that the harvesting effort E is a
function of time t. For optimization, we use the concept of generalized Legendre condition. The
trajectory for the optimal singular extremal is derived and the optimal singular control is
determined. Lastly, one numerical example is taken up and graph for steady state is drawn to
illustrate the results. 8:52– 9:05 Morning Aug. 13 Room A Paper No: 1835
Lazy Learning for Multi-Class Classification using Genetic Programming
Hajira Jabeen1 and Abdul Rauf Baig2
1Iqra University
2National University of Computer and Emerging Sciences, Islamabad
In this paper we have proposed a lazy learning mechanism for multiclass classification using
genetic programming. This method is an improvement of traditional binary decomposition
method for multiclass classification. We train classifiers for individual classes for a certain
number of generations. Individual trained classifiers for each class are combined in a single
chromosome. A population of such chromosomes is created and evolved further. This method
suppresses the conflicting situations common in binary decomposition method. The proposed
lazy learning method has performed better than traditional binary decomposition method over
five benchmark datasets taken from UCI ML repository. 9:05– 9:18
A New hybrid algorithm for the multidimensional knapsack problem
40
Morning Aug. 13 Room A Paper No: 1585
Xiaoxia zhang, Zhe Liu and Qiuying Bai
College of Software Engineering, University of Science and Technology Liaoning,
China
This paper presents a novel hybrid algorithm to solve the multidimensional knapsack problem.
The main feature of this hybrid algorithm is to combine the solution construction mechanism of
ant colony optimization (ACO) into scatter search (SS). It considers both solution quality and
diversification. A new mechanism of the subset combination method has been applied
simultaneity, which hybridizes mechanism of the pheromone trail updating with combination
mechanism of scatter search to generate new solutions. Second, an improvement algorithm
should be embedded into the scatter search framework to improve solutions. Finally, the
experimental results have shown that our proposed method is competitive to solve the
multidimensional knapsack problem compared with the other heuristic methods in terms of
solution quality 9:18 – 9:31 Morning Aug. 13 Room A Paper No: 1905
Recursive and Incremental Learning GA Featuring Problem-Dependent Rule-Set
Haofan Zhang1, Lei Fang2 and Sheng-Uei Guan1
1 Department of Computer Science and Software Engineering, Xi’an Jiaotong-Liverpool
University, Suzhou, China;
2 School of Computer Science, University of St Andrews, St Andrews, UK
Traditional rule-based classifiers training with Genetic Algorithms have their major weaknesses
in the classification accuracy and training time. To resolve these drawbacks, this paper reviews
Recursive Learning of Genetic Algorithm with Task Decomposition and Varied Rule Set
(RLGA) and proposes its variation that features Incremental Attribute Learning (RLGA-I).
Experiments show that both the proposed solutions dramatically reduce the training duration
with better generalization accuracy. 9:31 – 9:45 Morning Aug. 13 Room A Paper No: 1663
Evolution of Product Kernels for Regularization Networks
Petra Vidnerova and Roman Neruda
Institute of Computer Science, Academy of Sciences of the Czech Republic,Pod
vodarenskou věži 2, Praha 8, Czech Republic
Approximation problems formulated as regularized minimization problems with kernel-based
stabilizers lead to solutions of the shape of linear combination of kernel functions. These can be
expressed as one-hidden layer feed-forward neural network schemes, however, the rich
possibilities of theoretical approach are usually not exploited in suitable learning algorithms. In
this paper we focus on regularization networks with product kernels and propose an
evolutionary learning algorithm utilizing genetic search for suitable parameters. The approach
is experimentally tested on experiments. 9:45– 10:00 Morning Aug. 13 Room A Paper No:
Evaluation of Crossover Operator Performance in Genetic Algorithms with Binary
Representation
Stjepan Picek, Marin Golub, and Domagoj Jakobovic
Faculty of Electrical Engineering and Computing, Unska 3, Zagreb, Croatia
Genetic algorithms (GAs) generate solutions to optimization problems using techniques
41
1906 inspired by natural evolution, like crossover, selection and mutation. In that process, crossover
operator plays an important role as an analogue to reproduction in biological sense. During the
last decades, a number of di_erent crossover operators have been successfully designed.
However, systematic comparison of those operators is di_cult to _nd. This paper presents a
comparison of 10 crossover operators that are used in genetic algorithms with binary
representation. To achieve this, experiments are conducted on a set of 15 optimization
problems. A thorough statistical analysis is performed on the results of those experiments. The
results show signi_cant statistical di_erences between operators and an overall good
performance of uniform, single-point and reduced surrogate crossover. Additionally, our
experiments have shown that orthogonal crossover operators perform much poorer on the given
problem set and constraints.
10:00 – 10:10
Room A Coffee Break
Swarm Intelligence and Optimization Chair Mehran Yazdi, Yanxia Sun 10:10 – 10:23 Morning Aug. 13 Room A Paper No: 1624
Immune Gravitation inspired Optimization Algorithm
Yu Zhang, Lihua Wu, Ying Zhang and Jianxin Wang
College of Information Science and Technology, Hainan Normal University,
The traditional Gravitational Search Algorithm (GSA) has the advantages of easy
implementation, fast convergence and low computational cost. However, GSA driven by the
gravity law is easy to fall into local optimum solution. The convergence speed slows down in
the later search stage, and the solution precision is not good. Inspired by the biological immune
system, we introduce the characteristics of antibody diversity and vaccination, and propose a
novel immune gravitation optimization algorithm (IGOA) to help speed the convergence of
evolutionary algorithms and improve the optimization capability. The comparison experiments
of IGOA, GSA and PSO on some benchmark functions are carried out. The proposed algorithm
shows competitive results with improved diversity and convergence. It provides new
opportunities for solving previously intractable function optimization problems. 10:23– 10:36 Morning Aug. 13 Room A Paper No: 1258
Stem Cells Optimization Algorithm
Mohammad Taherdangkoo1 , Mehran Yazdi1 and Mohammad Hadi Bagheri2
1 Department of Communications and Electronics, Faculty of Electrical and Computer
Engineering, Shiraz University, Shiraz, Iran
2 Radiology Medical Imaging Research Center, Shiraz University of Medical Sciences,
Shiraz, Iran
Optimization algorithms have been proved to be good solutions for many practical applications.
They were mainly inspired by natural evolutions. However, they are still faced to some
problems such as trapping in local minimums, having low speed of convergence, and also
having high order of complexity for implementation. In this paper, we introduce a new
optimization algorithm, we called it Stem Cells Algorithm (SCA), which is based on behavior
42
of stem cells in reproducing themselves. SCA has high speed of convergence, low level of
complexity with easy implementation process. It also avoid the local minimums in an intelligent
manner. The comparative results on a series of benchmark functions using the proposed
algorithm related to other well-known optimization algorithms such as genetic algorithm (GA),
particle swarm optimization (PSO) algorithm, ant colony optimization (ACO) algorithm and
artificial bee colony (ABC) algorithm demonstrate the superior performance of the new
optimization algorithm. 10:36 – 10:49 Morning Aug. 13 Room A Paper No: 1302
A Population-based Hybrid Extremal Optimization Algorithm
Yu Chen, Kai Zhang, and Xiufen Zou�
School of Mathematics and Statistics, Wuhan University, Wuhan, 430072, China.
The extremal optimization (EO) algorithm is a kind of evolutionary algorithm which has been
applied successfully in combinatorial optimization, while its application on continuous
optimization encounters the problems of heavy complexity and weak exploration ability. This
paper proposes a new hybrid population-based EO algorithm, named as the adaptive
co-evolution population-based extremal optimization (ACPEO) algorithm, in which all
individuals co-evolve adaptively with each other and the differential evolution (DE) operator is
incorporated to improve the global search ability. By employing a novel evaluation method of
variables, the ACPEO algorithm performs well on several kind of benchmark problems.
Experimental results show that the ACPEO algorithm is robust due to the capability for solving
different problems with the same parameter setting, and it is also stable because changes in the
parameters’ values do not influence its performances seriously. 10:49– 11:02 Morning Aug. 13 Room A Paper No: 1575
An Effective Ant Colony Algorithm for Graph Planarization Problem
Li-Qing Zhao1, Cui Zhang2, and Rong-Long Wang1
1Graduate School of Engineering, University of Fukui,Bunkyo 3-9-1, Fukui-shi, Japan
2Department of Autocontrol, Liaoning Institute of Science and Technology, Benxi,
China
In this paper, an effective ant colony algorithm is proposed for solving the graph planarization
problem. In the proposed algorithm, two kinds of pheromone are adopted to reinforce the search
ability, and each kind of pheromone consists of two elements. The proposed algorithm is
verified by a large number of simulation runs and compared with other algorithms. The
experiment results show that the proposed algorithm performs remarkably well and outperforms
its competitors. 11:02– 11:15 Morning Aug. 13 Room A Paper No: 1689
A Novel Chaos Glowworm Swarm Optimization Algorithm for Optimization Functions
Kai Huang1 and Yong quan Zhou1,2
1College of Mathematics and Computer Science Guangxi University for Nationalities
Nanning, Guangxi 530006, China
2Guangxi Key Laboratory of Hybrid Computation and IC Design Analysis, Nanning
Guangxi 530006, China
This paper a novel chaotic glowworm swarm optimization algorithm (CGSO) is proposed. In
CGSO algorithm, the chaotic search strategies are incorporated in GSO to initialize the first
iteration solutions, so that it can obtain high-quality and evenly distributed initial solutions, and
43
avoids GSO being trapped in local optima, each glowworm disturbs by chaos in a disturbance
range can get more precise global solution. Compared with GSO algorithm, experiments with
six test functions shows that convergence quality and precision are improved, which testify that
CGSO are valid and feasible. 11:15 – 11:28 Morning Aug. 13 Room A Paper No: 1565
Fully Connected Multi-objective Particle Swarm
Zenghui Wang1 and Yanxia Sun2
1School of Engineering, University of South Africa, Florida 1710, South Africa
2Department of Electrical Engineering &French South African Institute of Technology
(F'SATI), Tshwane University of Technology, Pretoria 0001, South Africa
In this paper, a new model for multi-objective particle swarm optimization (MOPSO) is
proposed. In this model, each particle's behavior is influenced by the best experience among its
neighbors, its own best experience and all its components. The influence among different
components of particles is implemented by the on-line training of a multi-input Multi-output
back propagation (BP) neural network. The inputs and outputs of the BP neural network are the
particle position and it’s the 'gradient descent' direction vector to the less objective value
according to the definition of no-domination, respectively. Therefore, the new structured
MOPSO model is called a fully connected multi-objective particle swarm optimizer
(FCMOPSO). Simulation results and comparisons with exiting MOPSOs demonstrate that the
proposed FCMOPSO is more stable and can improve the optimization performance. 11:28– 11:41 Morning Aug. 13 Room A Paper No: 1348
A New Multi-swarm Multi-objective Particle Swarm Optimization based on Pareto Front
Set
Yanxia Sun1, Barend Jacobus van Wyk1 and Zenghui Wang2
1 French South African Institute of Technology (F'SATI), Tshwane University of
Technology, Pretoria 0001, South Africa.
2 School of Engineering, University of South Africa, Pretoria 0003, South Africa
In this paper, a new multi-swarm method is proposed for multi-objective particle swarm
optimization. To enhance the Pareto front searching ability of PSO, the particles are divided into
many swarms. Several swarms are dynamically searching the objective space around some
points of the Pareto front set. The rest of particles are searching the space keeping away from
the Pareto front to improve the global search ability. Simulation results and comparisons with
existing Multi-objective Particle Swarm Optimization methods demonstrate that the proposed
method effectively enhances the search efficiency and improves the search quality. 11:41– 11:54 Morning Aug. 13 Room A Paper No: 1273
Constrained Clustering Via Swarm Intelligence
Xiaohua Xu1, Zhoujin
Pan1, Ping He2, Ling Chen1
1Department of Computer Science and Engineering, Yangzhou University, Yangzhou
225009, China
2Department of Computer Science and Engineering, Nanjing University of Aeronautics
and Astronautics, Nanjing 210016, China
This paper investigates the constrained clustering problem through swarm intelligence. We
present an ant clustering algorithm based on random walk to deal with the pairwise constrained
44
clustering problems. Our algorithm mimics the behaviors of the real-world ant colonies and
produces better clustering result on both synthetic and UCI datasets compared with the
unsupervised ant-based clustering algorithm and the cop-kmeans algorithm. 11:55– 12:10 Morning Aug. 13 Room A Paper No: 1865
Memetic Fitness Euclidean-distance Particle Swarm Optimization for Multi-modal
Optimization
J. J. Liang1, Bo Yang Qu1,2, Song Tao Ma1, and Ponnuthurai Nagaratnam Suganthan2
1School of Electrical Engineering, Zhengzhou University,Zhengzhou, 450001, China
2School of Electrical and Electronic Engineering, Nanyang Technological University,
Singapore 639798, Singapore
In recent decades, solving multi-modal optimization problem has attracted many researchers
attention in evolutionary computation community. Multi-modal optimization refers to locating
not only one optimum but also the entire set of optima in the search space. To locate multiple
optima in parallel, many niching techniques are proposed and incorporated into evolutionary
algorithms in literature. In this paper, a local search technique is proposed and integrated with
the existing Fitness Euclidean-distance Ratio PSO (FER-PSO) to enhance its fine search ability
or the ability to identify multiple optima. The algorithm is tested on 8 commonly used
benchmark functions and compared with the original FER-PSO as well as a number of
multi-modal optimization algorithms in literature. The experimental results suggest that the
proposed technique not only increases the probability of finding both global and local optima
but also speeds up the searching process to reduce the average number of function evaluations.
Room B Intelligent Computing Theory and Application I
Chair Shulin Wang, Xutang Zhang 8:00 – 8:13 Morning Aug.13 Room B Paper No: 1920
Dimensional Reduction based on Artificial Bee Colony for Classification Problems
Thananan Prasartvit1, Boonserm Kaewkamnerdpong2 and Tiranee Achalakul1
1Deprtment of Computer Engineering, 2Biological Engineering Program King
Mongkut’s University of Technology Thonburi, Thailand
High dimensionality of data is a limiting factor to data processing in many fields. It causes
ambiguousness in identifying significant factors for data analysis. Dimension reduction is
needed to separate irrelevant data from the desired data. This research proposes a novel method
for dimension reduction based on artificial bee colony (ABC). The method employs swarm
intelligence based on bee foraging model in order to select features that allow us to generate
subsets of dimensions from the original high-dimensional data while the resulting subsets
satisfy the defined objective. Support vector machine (SVM) is used in this study as fitness
evaluation of ABC in classification problems. To evaluate our method, we tested it with five
datasets and compared it with other dimension reduction algorithms. The result of this study
shows that using ABC and SVM is suitable for reducing the dimension of data. Moreover, this
approach provides efficient classification with high accuracy.
45
8:13-8:26 Morning Aug. 13 Room B Paper No: 1996
Network Security Situation Assessment Based on HMM
Boyun Zhang1,2,Zhigang Chen2,Shulin Wang3,Xiai Yan1,Dingxing Zhang4 and Qiang
Fan1
1 Department of Computer Science and Technology, Hunan Police Academy,
Changsha, Hunan 410138, China
2 School of Information Science and Engineering, Central South University, Changsha,
Hunan 410086, China
3School of Computer and Communication, Hunan University, Changsha, Hunan
410082, China
4 Computer Department, Guangdong Technical College of Water Resoureces &
Electric Engineering, Guangzhou, 510635, China
Network security situation assessment is the core of situation awareness, and is also a
qualitative and quantitative description of network security state. In this paper, the security state
and intrusion alarm event in the network or the host system have corresponded to the state and
observation symbols in HMM, so that a network security situation evaluation model based on
HMM has been proposed. This model intrudes the alarm sequences generated from the
detection system through association analysis to calculate the risk index of each host, so as to
give a quantitative evaluation of the risk status of the whole net-work system. The Network
Risk Index can be calculated easily and quickly. The experimental results show that this model
can effectively and accurately give a quantitative evaluation of the security state of the network
system. 8:26– 8:39 Morning Aug. 13 Room B Paper No: 1678
Closed-label Concept Lattice Based Rule Extraction Approach
Junhong Wang, Jiye Liang, and Yuhua Qian
School of Computer and Information Technology, Shanxi University, Taiyuan,
030006,Shanxi, China ; Key Laboratory of Computational Intelligence and Chinese
Information Processing of Ministry of Education, Taiyuan, 030006, China
Concept lattice is an e®ective tool for data analysis and extracting classi¯cation rules. However,
the classical concept lattice often produces a lot of redundant rules. Closed-label concept lattice
realizes reduction of concept intention, which can be used to extract fewer rules than the
classical concept lattice. This paper presents a method for classi¯cation rules extraction based
on the closed-label concept lattice. Examples show that the proposed method is e®ective for
extracting more concise classi¯cation rules. 8:39– 8:52 Morning Aug. 13 Room B Paper No: 1995
Network Security Situation Assessment Based on Hidden Semi-Markov Model
Boyun Zhang1,2, Zhigang Chen2, Xiai Yan1, Shulin Wang3 and Qiang Fan1
1 Department of Computer Science and Technology, Hunan Police Academy,
Changsha, Hunan 410138, China
2 School of Information Science and Engineering, Central South University, Changsha,
China
3 School of Computer and Communication, Hunan University, Changsha, 410082,
China
This paper takes use of the hidden semi-Markov model to evaluate network security situation.
46
HsMM modifies HMM model on the presumption that certain system status dwell time abides
with exponential distribution, which is more suitable to describe the actual situation of network
system operation.We propose the HsMM system status prediction algorithm under partial
observation conditions, and applies it into network security situation assessment. The
ex-periment result shows that HsMM could model system status dwell time, so it is very
propitious to make network system security assessment under complicated and changeable
attacks. 8:52– 9:05 Morning Aug. 13 Room B Paper No: 1244
Quantum Information Splitting Using GHZ-type and W-type States
Lvzhou Li1, Daowen Qiua2
1 Department of Computer Science, Sun Yat-sen University, Guangzhou 510006,
China,
2 SQIG–Instituto de Telecomunica¸c˜oes, IST, TULisbon, Av. Rovisco Pais 1049-001,
Lisbon, Portugal
Quantum information splitting (QIS) is such a procedure where a sender can distribute a
quantum information (state) to several recipients such that all of them can cooperate to recover
the quantum state, but any other subset can not. In this paper, we present a uniform method to
design QIS schemes using GHZ-type and W-type states. We also give a simple criteria on W
states for QIS. 9:05 – 9:18 Morning Aug. 13 Room B Paper No: 1998
Network Security Situation Assessment Based on Stochastic Game Model
Boyun Zhang1,2, Zhigang Chen2, Wensheng Tang4, Qiang Fan1,
1 Department of Computer Science and Technology, Hunan Police Academy,
Changsha, Hunan 410138, China
2 School of Information Science and Engineering, Central South University, Changsha,
Hunan 410086, China
3School of Computer and Communication, Hunan University, Changsha, 410082,
China
4Department of Computer Teaching, Hunan Normal
University,Changsha,410081,China
From the perspective of game theory, this paper studies the information security problem,
establishes an offense and defense game model of information security, and puts forward a
quantitative evaluation algorithm of network security based on stochastic game model. It makes
use of the network administrator’s evaluation of network equipment’s importance to determine
the game parameters, analyze Nash equilibrium and work out Nash strategy of the attacker and
the defender so as to obtain the probability distribution when network is in different secure
states and finally get the evaluation results of network security situation through quantitative
analysis. The game model of the attacker and the defender put forward in this paper provides a
new idea for solving information security problem in reality. 9:18 – 9:31 Morning Aug. 13 Room B
Forgeability Attack of Two Special Signature Schemes
Jianhong Zhang, Yuanbo Cui, Xi Wu
College of Science, North China University of Technology, Beijing 100144, P.R.China
Unforgeabilty is a primitive property of a secure digital signature. As two extensions of digital
47
Paper No: 1070
signature, proxy signature and certificateles blind signature play an important role in the
sensitive transmission. In this work, we analyze the security of two signature schemes, one is
the certificateless proxy blind signature scheme[6] which was proposed by Tso et al in
NSS2010, the other is an efficient proxy signature scheme[8] which was proposed by Hu et al
in NSWCTC 2010. Then, we show that the two schemes were insecure, meanwhile, we also
show that Tso et al's basic certificateless proxy signature is insecure. In Tso et al 's scheme, any
one can forge a signature on an arbitrary message. Finally, we give the corresponding attack
method and analyze the reason to produce such the attack, respectively. 9:31– 9:44 Morning Aug. 13 Room B Paper No: 1590
A Semantic Retrieval Framework for Engineering Domain Knowledge
Xutang Zhang1, Xiaofeng Chen1, Xin Hou2, Ting Zhuang1
1 School of Mechatronics Engineering, Harbin Institute of Technology Harbin, 150001,
Heilongjiang, China
2 Science and Technology on Electronmagnetic Scattering Laboratory, 100854, Beijing,
China
In this paper, we propose a knowledge retrieval framework based on semantically annotated
engineering ontology generated from domain documents. Particularly, we propose a scheme for
build relations between ontology and domain documents. First, we build anthologies in
engineering domain. Next, we transform the keywords into domain ontology concepts, and then
find the synonyms of these keywords which are used as real queries to directly input into the
query system. The semantic-based knowledge search and retrieval is then performed by
ontology mapping and comparison. Using the semantic network of ontology, this system not
only can conduct keyword-based retrieval, but also can understand the queries and answer
questions by fuzzy inference based on domain ontology. 9:45– 10:00 Morning Aug. 13 Room B Paper No: 1433
A Trojan Detector Generating Algorithm Based on Chaotic Theory
Jie Qin1, Qun Si1, Huijuan Yan1, Fuliang Yan1 1College of Information Science and Engineering, Henan University of Technology, Zhengzhou, China
After studying the existing detector generating algorithms used in the intrusion detection
systems, we improves the lacks of the algorithms and use them into Trojan detection system,
and propose a new approach of detector generating based on chaotic theory. The over-spread
character of chaos sequence combined the concept of weighted Euclidean distance was used to
generate set of detector with better distribution, and chaos initial value sensitivity was used to
enlarge the searching space. The experiment indicates that the algorithm not only remains the
diversity of population but also has fast astringency speed.. 10:00 – 10:10 Room B Coffee Break
Intelligent Computing Theory and Application II Chair Xibei Yang 10:10– 10:22
Mean-entropy Model for Portfolio Selection with Type-2 Fuzzy Returns
48
Morning Aug. 13 Room B Paper No: 1493
Ying Liu and Yanju Chen
College of Mathematics & Computer Science, Hebei University Baoding 071002, Hebei,
China.
Entropy is a measurement of the degree of uncertainty. Mean-entropy method can be used for
modeling the choice among uncertain outcomes. In this paper, we consider the portfolio
selection problem under the assumption that security returns are characterized by type-2 fuzzy
variables. Since the expectation and entropy of type-2 fuzzy variables haven't been well de¯ned,
type-2 fuzzy variables need to be reduced ¯rstly. Then we propose a mean-entropy model with
reduced variables. To solve the proposed model, we use the entropy formula of reduced fuzzy
variable and transform the mean-entropy model to its equivalent parametric form, which can be
solved by standard optimization solver. 10:22 – 10:34 Morning Aug. 13 Room B Paper No: 1855
Reducing Grammar Errors for Translated English Sentences
Nay Yee Lin1, Khin Mar Soe2, and Ni Lar Thein3
1,3University of Computer Studies, Yangon
2Natural Language Processing Laboratory;University of Computer Studies, Yangon
One challenge of Myanmar-English statistical machine translation system is that the output
(translated English sentence) can often be ungrammatical. To address this issue, this paper
presents an ongoing grammar checker as a second language by using trigram language model
and rule based model. It is able to solve distortion, deficiency and make smooth the translated
English sentences. We identify the sentences with chunk types and generate context free
grammar (CFG) rules for recognizing grammatical relations of chunks. There are three main
tasks to reduce grammar errors: detecting the sentence patterns in chunk level, analyzing the
chunk errors and correcting the errors. Such a three level scheme is a useful framework for a
chunk based grammar checker. Experimental results show that the proposed grammar checker
can improve the correctness of translated English sentences. 10:34– 10:46 Morning Aug. 13 Room B Paper No: 1444
The Models of Dominance–based Multigranulation Rough Sets
Xibei Yang
Jiangsu Sunboon Information Technology Co., Ltd., Wuxi, Jiangsu, 214072, P.R. China
In this paper, the dominance–based rough set approach is introduced into multigranulation
environment. Two different dominance–based multigranulation rough sets models:
dominance–based optimistic multigranulation rough set and dominance–based pessimistic
multigranulation rough set are constructed, respectively. Not only the properties of these two
dominance–based multigranulation rough sets are discussed, but also the relationships among
dominance–based optimistic multigranulation rough set, dominance–based pessimistic
multigranulation rough set and the classical dominance–based rough set are investigated 10:46– 10:58 Morning Aug. 13 Room B Paper No: 1445
An Intuitionistic Fuzzy Dominance–based Rough Set
Yanqin Zhang1 and Xibei Yang2
1 School of Economics, Xuzhou Institute of Technology, Xuzhou, 221000, P.R. China
2 School of Computer Science and Engineering, Jiangsu University of Science and
Technology, Zhenjiang, Jiangsu, 212003, P.R. China
49
The dominance–based rough set approach plays an important role in the development of the
rough set theory. It can be used to describe the inconsistencies coming from consideration of the
preference–ordered domains of the attributes. The purpose of this paper is to further generalize
the dominance–based rough set model to fuzzy environment. The constructive approach is used
to define the intuitionistic fuzzy dominance–based lower and upper approximations,
respectively. Basic properties of the intuitionistic fuzzy dominance–based rough
approximations are then examined. 10:58 – 11:10 Morning Aug. 13 Room B Paper No: 1474
A Covering-based Pessimistic Multigranulation Rough Set
Guoping Lin and Jinjin Li
Department of Mathematics and Information Science, Zhangzhou Normal University,
Zhangzhou, 363000, Fujian, China
In view of granular computing, the classical optimistic and pessimistic multigranulation rough
set models are both primarily based on simple granules among multiple granular structures,
namely multiple partitions of the universe in MGRS. This correspondence paper
presents a new rough set model where set approximations are de ned by using multiple
coverings on the universe. In order to distinguish Qian's covering-based optimistic
multigranulation rough set model, we call the new rough set model as covering-based
pessimistic multigranulation rough set model. The key distinction between covering-based
pessimistic multigranulation rough set model and Qian's covering-based optimistic
multigranulation rough set model is set approximation descriptions. Then some properties are
proposed for covering-based pessimistic multigranulation rough set model. 11:10 – 11:22 Morning Aug. 13 Room B Paper No: 1486
A generalized multi-granulation rough set approach
Weihua Xu, Xiantao Zhang and Qiaorong Wang
School of Mathematics and Statistics,Chongqing University of Technology, 400054
Chongqing, P.R. China
A generalized multi-granulation rough set is proposed in thispaper. In the new model,
supporting characteristic function is de ned and a parameter called information level is
introduced to investigate that an object supports a concept precisely under majority
granulations.Moreover, some important properties are discussed on the new multi-granulation
rough set. And it can be found that the proposed model is more valid than old multiple
granulation rough set models and Pawlak rough set model. 11:22– 11:34 Morning Aug. 13 Room B Paper No: 1098
A Text Classication Algorithm Based on Rocchio and Hierarchical Clustering
Anping Zeng1 and Yongping Huang2 1School of Computer and Information Engineering, Yibin University,Yibin, Sichuan
644007, China, 2Computational Physics Key Laboratory of Sichuan Province, Yibin
University,Yibin, Sichuan 644007, China
The disadvantages of traditional classification algorithms are firstly discussed. Then, a new
algorithm called HI-Rocchio is proposed. The HI-Rocchio algorithm includes two parts. The
first part is an incremental Rocchio algorithm based on Rocchio algorithm, and the second is an
improved Hierarchical clustering algorithm. The HI-Rocchio algorithm not only can generate
new classes incrementally but also get multi hierarchical relationship between classes.
50
Experiments verified the effectiveness of the algorithm 11:34 – 11:46 Morning Aug. 13 Room B Paper No: 1516
A Variable Muitlgranulation Rough Sets Approach
Ming Zhang1,2, Zhenmin Tang1, Weiyan Xu3, Xibei yang2
1 School of Computer Science and Technology, Nanjing University of Science and
Technology, Nanjing, Jiangsu, 210094, P.R. China
2 School of Computer Science and Engineering, Jiangsu University of Science and
Technology, Zhenjiang, Jiangsu, 212003, P.R. China
3 School of Mathematics and Physics, Jiangsu University of Science and Technology,
Zhenjiang, Jiangsu, 212003, P.R. China
By analyzing the limitations of optimistic multigranulation rough set and pessimistic
multigranulation rough set, the concept of the variable multigranulation rough set is proposed.
Such multigranulation rough set is a generalization of both optimistic and pessimistic
multigranulation rough set. Furthermore, not only the basic properties about the variable
multigranulation rough set is discussed, but also the relationships among optimistic, pessimistic
and variable multigranulation rough sets are deeply explored. These results are meaningful for
the development of multigranulation rough set theory. 11:46 – 11:58 Morning Aug. 13 Room B Paper No: 1763
Incomplete multigranulation rough sets in incomplete ordered decision system
Li‐juan Wang 1,2 ;Xi‐bei Yang1,2; Jing‐yu Yang 1 ;Chen Wu 2
1 School of Computer Science and Technology, NUST, Nanjing 210094, China
2 Department of Computer Science and Engineering, JUST, Zhengjiang 212003, China
The tolerance relation based incomplete multigranulation rough set is not able to explore the
incomplete ordered decision systems. To solve such problem, similarity dominance relation
based rough set approach is introduced into multigranulation environment in this paper. Two
different types of models: similarity dominance relation based optimistic incomplete
multigranulation rough set model are constructed respectively. The properties and the
relationships of them are discussed. Eight types of decision rules in the two models are
proposed. An illustrative example is employed.
11:58– 12:10 Morning Aug. 13 Room B Paper No: 1245
Attacks and improvements of QSDC schemes based on CSS Codes
Xiangfu Zou1,2 and Daowen Qiu1,3
1 Department of Computer Science, Sun Yat-sen University, Guangzhou 510006,
China
2 School of Mathematics and Computational Science, Wuyi University, Jiangmen
529020, China
3 SQIG–Instituto de Telecomunica¸c˜oes, IST, TULisbon, Av. Rovisco Pais 1049-001,
Lisbon, Portugal
Quantum secure direct communication (QSDC) is a new research direction in quantum
cryptography. Recently, a QSDC scheme using quantum Calderbank-Shor-Steane (CSS) error
correction codes is proposed [Journal of Software, 17, 509–515 (2006)]. However, in this paper,
it is showed that the scheme can not resist the man-in-the-middle attack. To resist
man-in-the-middle attacks, an improved QSDC scheme using quantum CSS error correction
51
codes is suggested. Furthermore, we discuss the security of the improved QSDC scheme for an
ideal noiseless channel. In particular, it can resist the attack without eavesdropping.
Room C Sino-Korean Intelligent Information Processing and Automation
Workshop I Chair De-shuang Huang, Kang-Hyun Jo 8:00– 8:13 Morning Aug. 13 Room C Paper No: 1156
Implementation of Interactive Interview System using Hand Gesture Recognition
Yang Weon Lee
Department of Information and Communication Engineering, Honam University,
Seobongdong, Gwangsangu, Gwangju, 506-714, South Korea
This paper describes a implementation of virtual interactive interview system. A hand motion
recognition algorithm based on the particle lters is applied for this system. The particle lter is
well operated for human hand motion recognition than any other recognition algorithm.
Through the experiments, we show that the proposed scheme is stable and works well in virtual
interview system's environments. 8:13 – 8:26 Morning Aug. 13 Room C Paper No: 1588
Fuzzy PI Controller for Grid-Connected Inverters
Ngoc-Tung Nguyen*, Hong-Hee Lee
School of Electrical Engineering, University of Ulsan, Nam-gu, Ulsan, Korea
In this paper, a current controller for grid-connected inverter is proposed by using a fuzzy logic
control algorithm. In PI controller to control the grid-connected inverter, the gains of PI
controller are changed with the aid of the fuzzy logic algorithm in order to get the fast transient
performance despite of the input variations and load disturbances. The inputs of fuzzy logic
controller are the error between the measured currents and the reference values in rotating
reference frame, and the derivation of regulated voltage. The effectiveness of proposed
controller strategy has been verified by simulation with PSIM software and compared with that
of the conventional PI controller. 8:26 – 8:39 Morning Aug. 13 Room C Paper No: 1866
A Comprehensive Study on IEC61850 Process Bus Architecture and Spit Bus based
Differential Protection
Mojaharul Islam* and Hong-Hee Lee**
School of Electrical Engineering, University of Ulsan, Nam-gu, Ulsan, Korea
IEC61850 communication standard for digital substation automation creates a new way to think
about conventional protection scheme and configuration of substation. The presence of
communication link in process bus makes a revolutionary change for future digital substation.
This paper discusses briefly about process bus architecture and presents a new approach of
thinking in busbar structure to make the bus protection less sensitive with communication link
52
performance degradation. IEC61850-9-2LE implementation is analyzed and process bus based
on Ethernet technology is considered. Transient element based conventional bus differential
protection is used with split busbar configuration and finally demonstrates the reliability and
feasibility of proposed technique for digital substation. 8:39– 8:52 Morning Aug. 13 Room C Paper No: 1256
Korean Documents Copy Detection based on Ferret
Byung Ryul Ahn1, Won-gyum Kim2, Won Young Yu3, Moon-Hyun Kim1
1Artificial Intelligence Lab, School of Computer Engineering, SungKyunKwan Univ., 300 Chunchun-Dong, Jangan-Ku, Suwon-si 440-746, South Korea 2Copyright Protection Center, Sangam-Dong, Mapo-Gu, Seoul 121-270, South Korea 3Contents Research Division, ETRI, 161 Gajeong-dong, Yuseong-gu, Daejon, South Korea
With the development of electronic documents, plagiarism is rapidly increasing and, given the
difficulty of manual detection, need for plagiarism detection systems to help protect intellectual
property has emerged. Many content-based detection systems have been developed and are
actually used in some foreign countries, but they are still insufficient for documents in Korean.
In particular, the high variance of Hangul makes the development of detection systems more
difficult. This study proposes a Hangul document detection method based on Ferret’s trigrams.
Ferret only considered the frequency of trigram matches as a way to detect similarity, but in this
study the system is developed further by weighting results depending on the degree of trigram
match, thereby improving the accuracy of similarity detection. 8:52 – 9:05 Morning Aug. 13 Room C Paper No: 1124
An Improved Newman Algorithm for MiningAn Improved Newman Algorithm for
Mining Overlapping Modules from Protein-Protein Interaction Networks
Xuesong Wang, Lijing Li, Yuhu Cheng
School of Information and Electrical Engineering, China University of Mining and
Technology, Xuzhou, Jiangsu 221116, P.R.China
With the development of high-throughput technologies in recent years, more and more
scientists focus on protein-protein interaction (PPI) networks. Previous studies showed that
there are modular structures in PPI networks. It is well known that Newman algorithm is a
classical method for mining associations existed in complex networks, which has advantages of
high accuracy and low complexity. Based on the Newman algorithm, we proposed an improved
Newman algorithm to mine overlapping modules from PPI networks. Our method mainly
consists of two steps. Firstly, we try to discover all candidate nodes whose neighbors belong to
more than one module. Secondly, we determine candidate nodes that have positive effects on
modularity as overlapping nodes and copy these nodes into their corresponding modules. In
addition, owing to the features of existing system noise in PPI networks, we designed
corresponding methods for de-noising. Experimental results concerning MIPS dataset show
that, the proposed improved Newman algorithm not only has the ability of finding overlapping
modular structure but also has low computational complexity. 9:05 – 9:18 Morning
Actor-Critic Algorithm Based on Incremental Least-Squares Temporal Difference with
Eligibility Trace
53
Aug. 13 Room C Paper No: 1126
Yuhu Cheng, Huanting Feng and Xuesong Wang
China University of Mining and Technology
Compared with value-function-based reinforcement learning (RL) methods, policy gradient
reinforcement learning methods have better convergence, but large variance of policy gradient
estimation influences the learning performance. In order to improve the convergence speed of
policy gradient RL methods and the precision of gradient estimation, a kind of Actor-Critic
(AC) learning algorithm based on incremental least-squares temporal difference with eligibility
trace (iLSTD(λ)) is proposed by making use of the characteristics of AC framework, function
approximator and iLSTD(λ) algorithm. The Critic estimates the value-function according to the
iLSTD(λ) algorithm, and the Actor updates the policy parameter based on a regular gradient.
Simulation results concerning a grid world with 10×10 size illustrate that the AC algorithm
based on iLSTD(λ) not only has quick convergence speed but also has good gradient
estimation. 9:18– 9:31 Morning Aug. 13 Room C Paper No: 1241
Generating Test Data for both Paths Coverage and Faults Detection Using Genetic
Algorithms
Dun-wei Gong and Yan Zhang
School of Information and Electrical Engineering, China University of Mining and
Technology,
Various studies on generating test data have been done up to date, but few test data generated by
these studies can effectively detect faults lying in the program. We focus on the problem of
generating test data for both paths coverage and faults detection. First, the problem above is
formulated as a bi-objective optimization problem with one constraint, whose two objectives
are the number of faults detected in the traversed path and the risk level of these faults,
respectively, and the unique constraint is that the traversed path is just the target one; then, a
multi-objective evolutionary algorithm is employed to effectively solve the formulated model;
finally, the proposed method is applied in bubble sort program manually injected with some
faults, and compared with the random method and the evolutionary optimization one without
the task of detecting faults. The experimental results confirm the advantage of our method. 9:31– 9:45 Morning Aug. 13 Room C Paper No: 1325
Simulation of Visual Attention using Hierarchical Spiking Neural Networks
QingXiang Wu1,2, Martin McGinnity1, Liam Maguire1, Rongtai Cai2, Meigui Chen2
1Intelligent Systems Research Center, School of Computing and Intelligent Systems
University of Ulster at Magee, Londonderry, BT48 7JL, Northern Ireland, UK
2School of Physics and OptoElectronics Technology, Fujian Normal University Fuzhou,
350007, China
Based on the information processing functionalities of spiking neurons, a hierarchical spiking
neural network model is proposed to simulate visual attention. The network is constructed with
a conductance-based integrate-and-fire neuron model and a set of specific receptive fields in
different levels. The simulation algorithm and properties of the network are detailed in this
paper. Simulation results show that the network is able to perform visual attention to extract
objects based on specific image features. Using extraction of horizontal and vertical lines, a
demonstration shows how the network can detect a house in a visual image. Using this visual
54
attention principle, many other objects can be extracted by analogy. 9:45 – 10:00 Morning Aug. 13 Room C Paper No: 1400
Using Heart Rate Variability Parameter-based Feature Transformation Algorithm for
Driving Stress Recognition
Jeen-Shing Wang1, Che-Wei Lin1, and Ya-Ting C. Yang2
1Department of Electrical Engineering,
2Institute of Education & Center for Teacher Education, National Cheng Kung
University Tainan 701, Taiwan, R.O.C.
This paper presents a heart rate variability (HRV) parameter-based feature transformation
algorithm for driving stress recognition. The proposed parameter-based transformation
algorithm consists of feature generation, feature selection, and feature dimension reduction. In
order to generate significant features from ECG signals, parameter-based feature generation
method is proposed in this study. The parameter-based method calculates features from
five-minute HRV analysis. The kernel-based class separability (KBCS) is employed as the
selection criterion for feature selection. To reduce computational load of the algorithm,
principal component analysis (PCA) and linear discriminant analysis (LDA) are adopted for
feature dimension reduction. Our experimental results show that the combination of KBCS,
LDA, and PCA can achieve satisfactory recognition rates for the features generated by
parameter-based feature generation method. The main contribution of this study is that our
proposed approach can use only ECG signals to effectively recognize driving stress conditions
with very good recognition performance.
10:00 – 10:10
Room C Coffee Break
Sino-Korean Intelligent Information Processing and Automation
Workshop II Chair Kang-Hyun Jo, De-Shuang Huang 10:10 – 10:23 Morning Aug. 13 Room C Paper No: 1367
Comparison of Scalable ACC and MC-CDMA for Practical Video Fingerprinting Scheme
Liu Feng and Seong Whan Kim
School of Computer Science, University of Seoul, Seoul, Korea
Fingerprinting is used to determine originators of unauthorized copies. Multiple users may
collude by creating an average or median of their individual fingerprinted copies. Early
fingerprinting research including ACC (anti-collusion code) cannot support large number of
users. There have been two fingerprint researches for practically large user group support:
SACC (scalable ACC) scheme and MC-CDMA (multi-carrier code-division multi-access) based
fingerprinting scheme. In SACC scheme, they use a codebook extending ACC using Gaussian
distributed random variable and use angular decoding scheme for average, median, LCCA
attack robustness. MC-CDMA scheme uses direct spreading approach for identifying large
group of users. In this paper, we compare two schemes in three aspects: imaging quality,
55
computational complexity, and user capacity. Our experimental results show that SACC scheme
has achieved better imaging quality and lower computational complexity which are important
for video fingerprinting performance. MC-CDMA scheme outperforms SACC in user capacity
as MC-CDMA uses direct spreading based CDMA approach while SACC uses a frequency
hopping based CDMA approach. 10:23 – 10:36 Morning Aug. 13 Room C Paper No: 1413
A Method to Improve Performance of Heteroassociative Morphological Memories
Naiqin Feng1, Yushan Zhang1, Lianhui Ao1 and Shuangxi Wang1
1 College of Computer & Information Technology, Henan Normal University
General speaking, the heteroassociative morphological memory (HMM) is incomplete, namely,
it cannot give a guarantee of perfect recall memory, even though without any input noises. The
paper focuses on the problem and proposes a new method to improve performance of
heteroassociative morphological memories. This method can realize the perfect recall of HMMs
for perfect inputs or within a certain range of noises. An example is provided to illustrate the
proposed method and its performance. 10:36 – 10:49 Morning Aug. 13 Room C Paper No: 1618
High-Performance Video based Fire Detection Algorithms Using a Multi-Core
Architecture
Yongmin Kim1, Myeongsu Kang1, and Jong-Myon Kim1,
1 School of Electrical Engineering, University of Ulsan, Ulsan, South Korea
As fire accidents usually cause economic and environmental damage, including the loss of
human lives, video-based fire detection has become more appealing in surveillance systems.
However, video based fire detection algorithms demand tremendous computational and I/O
requirements. To meet these requirements, we introduce an SIMD (Single Instruction Multiple
Data) based multi-core architecture that consists of 16 processing elements (PEs) and small
local memory. In addition, we compare the performance and efficiency of the multi-core
architecture with a commercial Texas Instrument digital signal processor (TI DSP) to
demonstrate the potential for improved performance of the multi-core architecture.
Experimental results indicate that the multi-core architecture is 27.18 times and 3.89 times
better than TI DSP in terms of execution time and energy efficiency, respectively. 10:49 – 11:02 Morning Aug. 13 Room C Paper No: 1649
Building Face Reconstruction from Sparse View of Monocular Camera
My-Ha Le and Kang-Hyun Jo
Graduated School of Electrical Engineering, University of Ulsan, Ulsan, Korea
According to this method, building faces are detected by using color, straight line, edge and
vanishing point. In the next step, building faces from multi view are extracted. Point clouds of
building are obtained from triangulation step. The building faces are reconstructed by plane
fitting afterward. Texture mapping is applied to render the true surface. The simulation results
will demonstrate the effectiveness of this method. 11:02 – 11:15 Morning Aug. 13 Room C
Sparse Maximum Margin Discriminant Analysis for Gene Selection
Yan Cui1, Jian Yang1 and Chun-Hou Zheng2
1School of Computer Science and Technology, Nanjing University of Science and
Technology, Nanjing, Jiangsu, China.
2College of Information and Communication Technology, Qufu Normal University,
56
Paper No: 1475
Rizhao, Shandong, China.
Dimensionality reduction is necessary for gene expression data classification. In this paper,
based on sparse representation, we propose a sparse maximum margin discriminant analysis
(SMMDA) method for reducing the dimensionality of gene expression data. It could find the
one dimension projection in the most separable direction of gene expression data, thus one can
use sparse representation technique to regress the projection to obtain the relevance vector for
the gene set and select genes according to the vector. Extensive experiments on publicly
available gene expression datasets show that SMMDA is efficient for gene selection 11:15 – 11:28 Morning Aug. 13 Room C Paper No: 1491
Tracking Objects Using Orientation Covariance Matrices
Peihua Li and Qi Sun
School of Computer Science and Technology, Heilongjiang University, Harbin, China
This paper presents a novel model, called orientation covariance matrices, to represent the
object region and introduces a steepest descent method for object tracking. This model
partitions the gradient orientation space of joint color channels into subspaces (bins), and
computes the covariance matrix of image features in every bin. In the model a feature point
does not belong exclusively to one bin; instead, it makes contributions to several neighboring
bins. This is accomplished by introducing the cosine function for weighting the gradient
components of feature vectors. The weighting function helps to alleviate the e®ect of errors in
the computation of gradients induced by noise and illumination change. We also introduce a
spatial kernel for emphasizing the feature vectors which are nearer to the object center and for
excluding more background information. Based on the orientation covariance matrices,we
introduce a distance metric and develop a steepest descent algorithm for object tracking.
Experiments show that the proposed method has better performance than the traditional
covariance tracking method. 11:28 – 11:41 Morning Aug. 13 Room C Paper No: 1566
An iris recognition approach with SIFT descriptors
Xiaomin Liu1 and Peihua Li2
1 School of Information and Electronics, Jia Mu Si University, China
2 School of Computer Science and Technology, Heilongjiang University
In iris recognition systems how to represent texture pattern is an important issue. The paper
proposes a novel approach based on SIFT for feature representation of iris texture. This
approach partitions a normalized iris image into non-overlapping small sub-images and uses
SIFT descriptor for representing the characteristics of each sub-image.As such the iris texture
pattern is represented by an ordered-set of SIFT descriptors. This representation is very
distinctive and insensitive to illumination changes.In addition; it encodes the positional
information of iris texture pattern. For iris matching we use Bhattacharyya distance to measure
the dissimilarity between two SIFT descriptors. The ¯nal distance is a sum of the distances of
the corresponding pairs of SIFT descriptors in two iris images. The experimental results on
UBIRIS.v1 and UBIRIS.v2 show that proposed method has promising performance. 11:41– 11:55 Morning Aug. 13
An Improved Extreme Learning Machine Based on Particle Swarm Optimization
Fei Han1, Hai-Fen Yao1, Qing-Hua Ling2
1School of Computer Science and Telecommunication Engineering, Jiangsu University,
57
Room C Paper No: 1644
Zhenjiang, Jiangsu, 212013, China
2School of Computer Science and Engineering, Jiangsu University of Science and
Technology,
Zhenjiang, Jiangsu, 212003, China
Traditional ELM may require high number of hidden neurons and lead to ill-condition problem
due to the random determination of the input weights and hidden biases. In this paper, we use a
modified particle swarm optimization (PSO) algorithm to select the input weights and hidden
biases of single-hidden-layer feedforward neural networks (SLFN) and Moore–Penrose (MP)
generalized inverse to analytically determine the output weights. The modified PSO optimizes
the input weights and hidden biases according to not only the root mean squared error on
validation set but also the norm of the output weights. The proposed algorithm has better
generalization performance than other ELMs and its conditioning is also improved. 11:55 – 12:10 Morning Aug. 13 Room C Paper No: 1770
A Novel DE-ABC-based Hybrid Algorithm for Global Optimization
Li Li1, Fangmin Yao1,Lijing Tan2, Ben Niu1,3﹡, and Jun Xu3
1College of Management, Shenzhen University Shenzhen 518060, China
2Management School, Jinan University, Guangzhou 510632, China;3e-Business
Technology Institute, The University of Hongkong, Hongkong, China
A novel hybrid swarm intelligent algorithm DEABC, integrating differential evolution (DE) and
artificial bee colony (ABC) algorithm, is proposed in this paper. By using global information
obtained form DE population and bee colony, the exploration and exploitation abilities of
DEABC algorithm are balanced. The DE population uses the global best to generate offspring
every generation. The bee colony acquires the best individual after few generations. The
experiments are performed on six benchmark functions to compare the efficiencies of DE,
ABC, PSO and DEABC. The numerical results indicate the proposed algorithm outperforms
other algorithms in terms of accuracy and convergence speed.
Afternoon, Saturday, August 13
Room A Intelligent Computing in Computer Vision
Chair Tarik Veli Mumcu 13:40 – 13:55 Afternoon Aug. 13 Room A
Research on Dynamic Human Object Tracking Algorithm
Yongjian He1,2, Qiong Wu2, Shoupeng Feng2, Rongkun Zhou2, Yonghua Xing2 and Fei
Wang1
1 Institute of Artificial Intelligence and Robotics, Xi’an Jiaotong University, Xi’an, China,
2 Xi’an Communication Institute, Xi’an, China
58
Paper No: 1129
This paper studies the dynamic human object tracking problem. Under the condition of both of
the camera and the object being tracked simultaneously move, when the movement of the object
is too fast and the speeds of the two do not match, the tracking of the moving object will have
lag issues. This paper presents an improved particle-tracking method. The method, during the
tracking process, can reduce the number of particles online according to the actual tracking
situation, thereby reducing computation time, so that the computing speed can be adjusted in
real time according to the velocity of the being-tracked object to form the best match of the
speeds. Experimental results show that the improved algorithm well solves the lag problems of
the moving object being tracked and the tracking performance is significantly improved. 13:55 – 14:10 Afternoon Aug. 13 Room A Paper No: 1489
Histogram Based Color Object Classification by Multi-Class Support Vector Machine Tarik Veli MUMCU, Ibrahim ALISKAN, Kayhan GULEZ and Gurkan TUNA
Yildiz Technical University 34349 Istanbul, Turkey
Trakya University, Vocational College of Technical Sciences, Department of Computer
Programming, Edirne, Turkey
This work presents a histogram based color object classification by SVM for laboratory
automation. In the laboratory environment, existing problem is the classification of color
objects which is understood as blob like pictures by the system via a camera. This automated
system is located at hospitals, blood banks where we introduce the system different blood
samples for different research purposes. The blood samples for different research purposes are
realized with different colors of tube caps. These caps constitute the main problem here since
their images are often blob like pictures. The segmented, multi color cap pictures are
investigated in this paper by SVM for color object classification. To validate the performance of
the system with SVM method, its output also compared to the other classification methods. In
the future work different color spaces will be incorporated with SVM for better color
classification. 14:10 – 14:25 Afternoon Aug. 13 Room A Paper No: 1418
Pavement crack segmentation algorithm based on local optimal threshold of cracks
density distribution
Shengchun Wang and Wensheng Tang
Department of Computer Teaching, Hunan Normal University, Changsha, China
Asphalt pavement distress is very important for road maintenance and rehabilitation decisions.
The traditional manual pavement crack detection by human eyes is expensive, labor intensive,
time consuming, and subjective. Automatic pavement distress detection algorithms are
developed quickly in recent years. Segmentation is one of important step in automated
pavement crack detect system. In this paper, a new segmentation algorithm by multi-scale and
local optimum threshold is developed. The algorithm was shown to be more effective and
robust than conventional segmentation algorithms.
14:25– 14:40
Integrated Real-Time Vision-Based Preceding Vehicle Detection in Urban Roads
59
Afternoon Aug.13 Room A Paper No: 1439
Yanwen Chong1, Wu Chen2, Zhilin Li2, William H. K. Lam3 and Qingquan Li1
1State Key Laboratory for Information Engineering in Surveying , Mapping and Remote
Sensing , Wuhan University , 129 Luoyu Road , Wuhan 430079 ,China
2Department of Land Surveying and Geoinformatics, Hong Kong Polytechnic
University, Kowloon, Hong Kong, China.
3Department of Civil and Structural Engineering, Hong Kong Polytechnic University,
Kowloon, Hong Kong, China.
This paper presents a real-time algorithm for a vision-based preceding vehicle detection system.
The algorithm contains two main components: vehicle detection with various vehicle features,
and vehicle detection verification with dynamic tracking. Vehicle detection is achieved using
vehicle shadow features to define a region of interest (ROI). After utilizing methods such as
histogram equalization, ROI entropy and mean of edge image, the exact vehicle rear box is
determined. In the vehicle tracking process, the predicted box is verified and updated. Test
results demonstrate that the new system possesses good detection accuracy and can be
implemented in real-time operation. 14:40 – 14:55 Afternoon Aug. 13 Room A Paper No: 1582
A Gesture Recognition System Using One-Pass DP Method
Takashi Kuremoto*,1, Yasuhiro Kinoshita*, Liang-bing Feng*, Shun Watanabe*, Kunikazu
Kobayashi* and Masanao Obayashi* * Graduate School of Science and Engineering, * Yamaguchi University
An online gesture recognition system using a dynamical programming, One-Pass DP, is
proposed in this paper. Firstly, 8 directions of hand motions are extracted with skin color
analysis and optical flow calculation using a primary visual cortex model. Then, the patterns of
motion are used to compose 40 basic templates of gestures. At last, hand gestures are
recognized by the One-Pass DP algorithm. Experiments dealt with individual and compound
gestures were executed by online processing, the results confirmed the effectiveness of the
proposed system. 14:55 – 15:10 Afternoon Aug. 13 Room A Paper No: 1646
Linear Pose Estimation Algorithm Based on Quaternion
Yongjian He1,2, Caigui Jiang1, Chengwei Hu3, Jingmin Xin1 ,
1 Institute of Artificial Intelligence and Robotics, Xi’an Jiaotong University, Xi’an, China
2 Xi’an Communication Institute, Xi’an, China
3China Academy of Space Technology, Beijing, China
A novel linear camera pose estimation algorithm is presented using known 3D to 2D line
correspondences and point correspondences. The rotation parameters are represented by unit
quaternion. For n (n>=4) correspondences, we establish an equation system with 2n quadratic
equations in thirteen variables and apply the “relinearization” method to obtain the rotation
parameters and translation parameters simultaneously. We compare our algorithm with
Ansar’s NLL algorithm for line correspondences by some synthetic experiments. Our algorithm
performs better on the aspect of running time and accuracy of determined pose parameters.
Some real experiments are produced by 1 point-3 lines, 2 points-2 lines, 3 points- 1 line
correspondences. The projection of a 3D model is applied to estimate the performance of our
algorithm.
60
15:10 – 15:25 Afternoon Aug. 13 Room A Paper No: 1540
Mass Classification with Level Set Segmentation and Shape Analysis for Breast Cancer
Diagnosis Using Mammography
Xiaoming Liu, Xin Xu, Jun Liu and J. Tang
College of Computer Science and Technology, Wuhan University of Science and
Technology, Wuhan 430081, China
Masses are the typical signs of breast cancer. Correctly classifying mammographic masses as
malignant or benign can assist radiologists to diagnosis breast cancer and can reduce the
unnecessary biopsy without increasing false negatives. In this paper, we investigate the
classification of masses with level set segmentation and shape analysis. Based on the initial
contour guided by the radiologist, level set segmentation is used to deform the contour and
achieve the final segmentation. Shape features are extracted from the boundaries of segmented
regions. Linear discriminant analysis and support vector machine are investigated for
classification. A dataset consists of 292 ROIs from DDSM mammogram images were used for
experiments. The method based on Fourier descriptor of normalized accumulative angle
achieved a high accuracy of Az=0.8803. The experimental results show that Fourier descriptor
of normalized accumulative angle is an effective feature for the classification of masses in
mammogram. 15:25– 15:40 Afternoon Aug. 13 Room A Paper No: 1571
An Video Shot Segmentation Scheme Based on Adaptive Binary Searching and SIFT1
Xinghao Jiang1, 4, Tanfeng Sun1, 4, Jin Liu2, 3, Wensheng Zhang3 and Juan Chao1
1 School of Information Security Engineering Shanghai Jiao Tong University,
2 State Key Lab. of Software Engineering, School of Computer Wuhan University,
3 Key Lab. of Complex System & Intelligence Science, Institute of Automation Chinese
Academy of Science;
4 Key Lab. of Shanghai Information Security Management and Technology Research
A video shot segmentation scheme with dual-detection model is proposed. In the pre-detection
round, the Uneven Blocked differences are presented and used in Adaptive Binary Search
(ABS) to detect shot boundaries. In the re-detection round, the Scale Invariant Feature
Transform (SIFT) method is applied to exclude false detections. Experiments show that this
algorithm achieves well performances in detecting both abrupt and gradual boundaries.
15:40-15:50
Room A Coffee Break
Intelligent Computing in Image Processing Chair Vandana Dixit Kaushik, Satoshi Mori 15:50– 16:05 Afternoon Aug. 13 Room A Paper No:
Image Enhancement Algorithm for Ink-on-paper Fingerprint
Raghav Agrawal, Badrinath Srinivas, and Phalguni Gupta Department of Computer Science & Engineering Indian Institute of Technology Kanpur,
Kanpur - 208016, India
Quality of a scanned ink-on-paper fingerprint (or offline fingerprint) is generally poorer than
61
1654 that of online fingerprint and therefore, it requires an efficient enhancement method. This paper
presents an algorithm to enhance the offline fingerprints using adaptive techniques. The
algorithm handles certain problems that are specific to this type of fingerprints. It involves
fingerprint extraction, intra-image classification followed by adaptive region enhancement of
classified regions. It has been tested on IIT Kanpur database. The results show that the proposed
algorithm performs well in enhancing ink-on-paper fingerprints. 16:05 – 16:20 Afternoon Aug. 13 Room A Paper No: 1669
Age Estimation using Active Appearance Models and Ensemble of Classifiers with
Dissimilarity-based Classification
Sharad Kohli, Surya Prakash and Phalguni Gupta
Department of Computer Science and Engineering, Indian Institute of Technology
Kanpur Kanpur 208016, India
This paper proposes a novel technique that uses Active Appearance Models (AAMs) and
Ensemble of classifiers for age estimation. In this technique, features are extracted from face
images by AAMs and a global classifier is then used to obtain an idea about the age by
distinguishing between child/teen-hood and adulthood, before age estimation. This is done by
an ensemble containing various classifiers trained on multiple dissimilarities and thereby which
reduces misclassification error. Different aging functions are considered for the classified
images to estimate age more accurately. Experiments are performed on the publicly available
FG-NET database. The method is found to be a good age estimator. 16:20 – 16:35 Afternoon Aug. 13 Room A Paper No: 1698
Non-Sampling Contourlet Based “ Consistency Verification” Method of Image Fusion
Shi-Bin Xuan, Gao-Li Sang, Bo Zhao and Zeng-Guo Zheng
School of Mathematics and Computer Science. Guangxi University for nationalities,
nanning, China
The image fusion is one of the important research areas of data fusion. The wavelet-based
fusion is the most popular image fusion methods at present. To remove "block" appeared in the
results of wavelet-based fusion, the paper proposes a window-based "consistency verification"
which is used to deal with the fusion result image obtained through Non-sampling contourlet
transform (NSCT). Within the window scope, the algorithm takes the current coefficient as the
centre point of the window, and statistics the coefficients derived from each source image. Then
the weights are assigned according to each source image occupy the proportional in the
window .At last the value of the window center can be automatically determined by
re-weighting. Experimental results show proposed algorithm is feasible and effective. 16:35 – 16:50 Afternoon Aug. 13 Room A Paper No: 1296
Understanding the Meaning of Shape Description
Satoshi Mori, Yoshinori Kobayashi, and Yoshinori Kuno
Department of Information and Computer Sciences, Saitama University, Saitama city,
Saitama 338-8570, Japan
Service robots need to be able to recognize objects located in complex environments. Although
there has been recent progress in this area, it remains difficult for autonomous vision systems to
recognize objects in natural conditions. Thus we propose an interactive object recognition
system, which asks the user to verbally provide information about the objects that it cannot
recognize. However, humans may use various expressions to describe objects. Meanwhile, the
62
same verbal expression may indicate different meanings depending on the situation. Thus we
have examined human descriptions of object shapes through experiments using human
participants. This paper presents the findings from the experiments useful in designing
interactive object recognition systems. 16:50 – 17:05 Afternoon Aug. 13 Room A Paper No: 1564
Fast Single Image Super-resolution by Self-trained Filtering
Dalong Li and Steven Simske
Hewlett Packard Company {dalong.li,steven.simske}@hp.com
This paper introduces an algorithm to super-resolve an image based on a self-training filter
(STF). As in other methods, we first increase the resolution by interpolation. The interpolated
image has higher resolution, but is blurry because of the interpolation. Then, unlike other
methods, we simply filter this interpolated image to recover some missing high frequency
details by STF. The input image is first downsized at the same ratio used in super-resolution,
then upsized. The super-resolution filters are obtained by minimizing the mean square error
between the upsized image and the input image at different levels of the image pyramid. The
best STF is chosen as the one with minimal error in the training phase. We have shown that STF
is more effective than a generic unsharp mask filter. By combining interpolation and filtering,
we achieved competitive results when compared to support vector regression methods and the
kernel regression method. 17:05– 17:20 Afternoon Aug. 13 Room A Paper No: 1655
No-Reference Image Quality Assessment for Facial Images
Debalina Bhattacharjee, Surya Prakash, and Phalguni Gupta
Department of Computer Science and Engineering,Indian Institute of Technology
Kanpur,Kanpur-208016, India
Image quality assessment traditionally means the comparison of original image with its
distorted version using conventional methods like Mean Square Error (MSE) or Peak Signal to
Noise Ratio (PSNR).In case of Blind Quality Evaluation with no prior knowledge about the
image, a single parameter becomes insu cient to de ne the overall image quality. This paper
proposes a quality metric based on sharpness of the image, presence of noise, overall contrast
and luminance of the image and the detection of the eyes. Experimental results reveal that the
proposed metric has strong relevance with human quality perception. 17:20 – 17:35 Afternoon Aug. 13 Room A Paper No: 1305
Palmprint based Recognition System using Local Structure Tensor and Force Field
Transformation
Kamlesh Tiwari, Devendra Kumar Arya, and Phalguni Gupta
Department of Computer Science and Engineering, Indian Institute of Technology
Kanpur Kanpur 208016, INDIA
This paper presents an efficient palmprint based recognition system. In this system, the image is
divided into disjoint sub-images. For each sub-image, the dominant orientation pixels based on
the force eld transformation are identi ed. Structure tensor values of these dominant orientation
pixels of each sub-image are averaged to form tensor matrix for the sub-image. Eigen
decomposition of each tensor matrix is used to generate the feature matrix which is used to take
decision on matching. The system has been tested on IITK database. The experimental results
reveal the accuracy of 100% for the database.
63
17:35 – 17:50 Afternoon Aug. 13 Room A Paper No: 1308
Modified Geometric Hashing for Face Database Indexing
Vandana Dixit Kaushik1, Amit K. Gupta2, Umarani Jayaraman2 and Phalguni Gupta2
1 Department of Computer Science & Engineering, Hartcourt Butler Technological
Institute, Kanpur 208002, India
2 Department of Computer Science & Engineering, Indian Institute of Technology
Kanpur, Kanpur 208 016, India
This paper presents a modified geometric hashing technique to index the database of facial
images. The technique makes use of minimum amount of search space and memory to provide
best matches with high accuracy against a query image. Features are extracted using
Speeded-Up Robust Features (SURF) operator. To make these features invariant to translation,
rotation and scaling, a pre-processing technique consisting of mean centering, principal
components, rotation and normalization has been proposed. The proposed geometric hashing is
used to hash these features to index each facial image in the database. It has achieved more than
99% hit rate for top 4 best matches.
Room B Workshop on Intelligent Computing in Scheduling
Chair Ling Wang 13:40 – 13:55 Afternoon Aug. 13 Room B Paper No: 1078
An Effective Shuffled Frog Leaping Algorithm for Solving Hybrid Flow-shop Scheduling
Problem
Ye Xu, Ling Wang, Gang Zhou, and Shengyao Wang
Tsinghua National Laboratory for Information Science and Technology (TNList),
Department of Automation, Tsinghua University, Beijing, 100084, P.R. China
In this paper, an effective algorithm based on the shuffled frog leaping algorithm (SFLA) is
proposed to solve the hybrid flow-shop (HFS) scheduling problem, which is a strong NP-hard
combinational problem with very wide engineering background. By using a special encoding
scheme and combining SFLA based memetic search and Meta-Lamarckian local search
strategy, the exploration and exploitation abilities are enhanced and well balanced for solving
the HFS problems. Simulation results based on some typical problems and comparisons with
some existing genetic algorithm and differential evolution demonstrate that the proposed
algorithm is effective and robust in solving the HFS problem. 13:55– 14:10 Afternoon Aug. 13 Room B Paper No: 1151
A Hybrid Algorithm Based on Simplex Search and Differential Evolution for
Resource-constrained Project Scheduling Problem
Ling Wang, Ye Xu, and Chen Fang
Tsinghua National Laboratory for Information Science and Technology (TNList),
Department of Automation, Tsinghua University, Beijing, 100084, P.R. China
In this paper, an effective hybrid algorithm is proposed to solve the resource-constrained project
scheduling problem by merging Nelder-Mead (NM) simplex method and differential evolution
64
(DE). The individuals are encoded with the priority value based method and decoded by serial
schedule generate scheme (SGS). Moreover, a reasonable framework is proposed to hybridize
the simplex-based geometric search and the DE-based evolutionary search, and the simplex
search is modified to further improve the quality of solutions obtained by DE. By interactively
using the two methods with different mechanisms, the searching behavior can be enriched and
the exploration and exploitation abilities can be well balanced. Simulation results based on
some benchmarks demonstrate the effectiveness of the proposed hybrid algorithm. 14:10 – 14:25 Afternoon Aug. 13 Room B Paper No: 1155
A Differential Evolution Algorithm for Lot-streaming Flow Shop Scheduling Problem
Hongyan Sang1, 2, Liang Gao1 and Xinyu Li1
1State Key Lab. of Digital Manufacturing Equipment & Technology in Huazhong
University of Science & Technology, Wuhan, 430074, PR China
A differential evolution (DE) algorithm is proposed to minimize the total weighted tardiness
and earliness penalties for lot-streaming flow shop scheduling problems. In the proposed DE
algorithm, the largest position value (LPV) rule is used to convert a real-number DE vector to a
job permutation. The DE evolution is used to perform global exploitation, and a local search
procedure is used to enhance the exploration capability. Extensive computational simulations
and comparisons are provided, which demonstrate the effectiveness of the proposed DE
algorithm. 14:25 – 14:40 Afternoon Aug. 13 Room B Paper No: 1099
An Estimation of Distribution Algorithm for the Flexible Job-shop Scheduling Problem
Shengyao Wang, Ling Wang, Gang Zhou, and Ye Xu
Tsinghua National Laboratory for Information Science and Technology (TNList),
Department of Automation, Tsinghua University, Beijing, 100084, P.R. China
In this paper, an effective estimation of distribution algorithm (EDA) is proposed to solve the
flexible job-shop scheduling problem with the criterion to minimize the maximum completion
time (makespan). With the framework of the EDA, the probability model is built with the
superior population and the new individuals are generated based on probability model. In
addition, an updating mechanism of the probability model is proposed and a local search
strategy based on critical path is designed to enhance the exploitation ability. Finally, numerical
simulation is carried out based on the benchmark instances, and the comparisons with some
existing algorithms demonstrate the effectiveness of the proposed algorithm. 14:40 – 14:55 Afternoon Aug. 13 Room B Paper No: 1094
An Effective Artificial Bee Colony Algorithm for Multi-objective Flexible Job-shop
Scheduling Problem
Gang Zhou, Ling Wang ,Ye Xu, and Shengyao Wang
Tsinghua National Laboratory for Information Science and Technology (TNList),
Department of Automation, Tsinghua University, Beijing, 100084, P.R. China
In this paper, an effective artificial bee colony (ABC) algorithm is proposed to solve the
multi-objective flexible job-shop scheduling problem with the criteria to minimize the
maximum completion time, the total workload of machines and the workload of the critical
machine simultaneously. By using the effective decoding scheme, hybrid initialization strategy,
crossover and mutation operators for machine assignment and operation sequence, local search
based on critical path and population updating strategy, the exploration and exploitation abilities
65
of ABC algorithm are stressed and well balanced. Simulation results based on some widely used
benchmark instances and comparisons with some existing algorithms demonstrate the
effectiveness of the proposed ABC algorithm. 14:55– 15:10 Afternoon Aug. 13 Room B Paper No: 1215
Discrete Harmony Search Algorithm for the No wait Flow Shop Scheduling Problem with
Makespan Criterion
Kaizhou Gao, Shengxian Xie, Hua Jiang, and Junqing Li
Computer school of Liaocheng University, Liaocheng, China, 252059
This paper presents a discrete harmony search (DHS) algorithm for solving no-wait flow shop
scheduling problems with makespan criterion. Firstly, a harmony is represented as a discrete job
permutation and three heuristic methods are proposed to initialize the harmony memory.
Secondly, by dynamically regrouping mechanism, the harmony memory is divided into several
groups for sharing information reciprocally. Thirdly, to stress the balance between the global
exploration and local exploration, a variable neighborhood search algorithm is developed and
embedded in the DHS algorithm. Computational simulation results based on the well-known
benchmarks and statistical performance comparisons are provided. Computational results and
comparison show the effectiveness of the presented DHS algorithm in solving the no-wait flow
shop scheduling with makespan criterion. 15:10 – 15:25 Afternoon Aug. 13 Room B Paper No: 1556
Hybrid Differential Evolution Optimization for No-wait Flow-shop Scheduling with
Sequence-dependent Setup Times and Release Dates
Bin Qian, Hua-Bin Zhou, Rong Hu, and Feng-Hong Xiang
Department of Automation, Kunming University of Science and Technology, Kunming
650051, China
In this paper, a hybrid algorithm based on differential evolution (DE), namely HDE, is proposed
to minimize the total completion time criterion of the no-wait flow-shop scheduling problem
(NFSSP) with sequence-dependent setup times (SDSTs) and release dates (RDs), which is a
typical NP-hard combinatorial optimization problem with strong engineering background.
Firstly, to make DE suitable for solving flow-shop scheduling problem, a largest-order-value
(LOV) rule is used to convert the continuous values of individuals in DE to job permutations.
Secondly, a speed-up evaluation method is developed according to the property of the NFSSP
with SDSTs and RDs. Thirdly, after the DE-based exploration, a problem-dependent local
search is developed to emphasize exploitation. Due to the reasonable balance between
DE-based global search and problem-dependent local search as well as the utilization of the
speed-up evaluation, the NFSSP with SDSTs and RDs can be solved effectively and efficiently.
Simulation results and comparisons demonstrate the superiority of HDE in terms of searching
quality, robustness, and efficiency. 15:25 – 15:40 Afternoon Aug. 13 Room B Paper No: 1395
Two Techniques to Improve the NEH Algorithm for Flow-shop Scheduling Problems
Gengcheng Liu, Shiji Song and Cheng Wu
Department of Automation, Tsinghua University, Beijing, 100084, PR China
Flow-shop scheduling problem (FSP) has been widely investigated in the area of manufacturing
systems. Up to now, the NEH algorithm is the best heuristic approach to solve FSP. However, in
large-scale problems, it takes quite long time for the NEH algorithm to find an approximate
66
optimal solution. In this paper, two new techniques are proposed to improve the NEH
algorithm. Firstly, to reduce the running time, block properties are developed and introduced to
NEH algorithm. Secondly, to obtain solutions with smaller makespan, tie-break rules are
applied. Simulation results show that these two techniques perform well in improving the NEH
algorithm.
15:40-15:50
Room B Coffee Break
Workshop on Intelligent Computing in Scheduling Chair Ling Wang 17:50– 18:10 Afternoon Aug. 13 Room B Paper No: 1832
Minimizing the total flowtime flowshop with blocking using a Discrete Artificial Bee
Colony
Yu-Yan Han1, Jun-Hua Duan1* ,Yu-Jie Yang2 , Min-Zhang1 and Bao-Yun1
1 School of Computer Science, Liaocheng University, China; 252059
2 Network center, Liaocheng University, China; 252059
This paper considers a discrete artificial bee colony (DABC) algorithm for the blocking flow
shop (BFS) scheduling problem to minimize total flowtime. The DABC algorithm utilizes
discrete job permutations to represent food sources and applies discrete operators to generate
new food sources for the employed bees, onlookers and scouts. First, an initialization scheme
based on MME (combination of MinMax and NEH) heuristic is presented to construct an initial
population with a certain level of quality and diversity. Second, a local search based on the
insertion neighborhood is applied to onlooker stage to improve the algorithm’s local
exploitation ability. Third, a destruction-construction operator is employed to obtain solutions
for the scout bees. Computational simulations and comparisons show that the proposed
algorithm (DABC) is effective and efficient for the blocking flow shop scheduling problems
with total flowtime criterion. 16:10 – 16:30 Afternoon Aug. 13 Room B Paper No: 1587
Variable Neighborhood Search for Drilling
Yun-Chia Liang1 and Chia-Yun Tien2
1,2 No 135 Yuan-Tung Road, Chungli City,Taoyuan County, 320 Taiwan
Among all types of production environment, identical parallel machines are frequently used to
increase the manufacturing capacity of the drilling operation in Taiwan printed circuit board
(PCB) industries. So when a manager plans the production scheduling, multiple but
conflicting objectives are often considered. Unlike the single objective problem, the
multiple-objective version no longer looks for an individual optimal solution, but a Pareto front
consisting of a set of non-dominated solutions. The manager then can select one of the
alternatives from the set. For this matter, our research aims at applying a variable
neighborhood search (VNS) algorithm in the identical parallel machine scheduling problem
(IPMSP) with two conflicting objectives: makespan and total tardiness. In VNS, two
neighborhoods are defined – insert a job to a different position or swap two jobs in the
67
sequence. To save the computational expense, one of the neighborhoods is randomly selected
for the target solution which is also arbitrarily chosen from the current Pareto front. The
proposed VNS algorithm is tested on a set of real data collected from a leading PCB factory in
Taiwan and its performance is compared with well-known methods in the literature. The
computational results show that VNS outperforms all competing algorithms – SPGA, MOGA,
NSGA-II, SPEA-II, and MACO in terms of solution quality and computational time. 16:30– 16:50 Afternoon Aug. 13 Room B Paper No: 1515
Research on Vehicle Routing Problem with Stochastic Demand Based on Multi-objective
Method
Yanwei Zhao, Chuan Li, Jing-ling Zhang, Xingqiu Ren, Wei Ren 1.Key Laboratory of Special Equipment and Advanced Processing Technology Ministry of Education,Zhejiang Univ. of Tech. , ZheJiang HangZhou 310012,China 2.College of Computer Science & Technology, Zhejiang Univ. of Tech. , ZheJiang Hangzhou 310023, China
This paper was targeted at minimizing the expectation of traveling distance maximizing the
expectation of customers’ degree satisfaction, a multi-objective vehicle routing problem with
stochastic demand (VRPSD) model based on soft time window was proposed. In order to solve
the problem, a hybrid PSO algorithm based on Pareto optimization method was designed in this
paper. The paper made the standard PSO algorithm discrete by re-defining operators and
employing swap recon, utilized challenge tournament method to construct Pareto optimal
solution set, applied an external archive to keep the diversity of solutions. Ultimately, a standard
example is used to verify the validity of the algorithm. 16:50 – 17:10 Afternoon Aug. 13 Room B Paper No: 1220
A Novel Multi-objective Particle Swarm Optimization Algorithm for Flow Shop
Scheduling Problems
Wanliang Wang1, Lili Chen2, Jing Jie1, Yanwei Zhao3, and Jing Zhang2
1 College of Computer Science and Technology, Zhejiang University of Technology,
Hangzhou 310023, China
2 College of Information Engineering, Zhejiang University of Technology, Hangzhou
310014, China
3 College of Mechanical Engineering, Zhejiang University of Technology, Hangzhou
310014, China
In this paper, a novel hybrid multi-objective particle swarm algorithm Mopsocd_BL is proposed
to solve the flow shop scheduling problem with two objectives of minimizing makespan and the
total idle time of machines. This algorithm bases on Baldwinian learning mechanism to
improve local search ability of particle swarm optimization, and uses the Pareto dominance and
crowding distance to update the solutions. Experimental results show that this algorithm can
maintain the diversity of solutions and find more uniformly distributed Pareto optimal solutions.17:10– 17:30 Afternoon Aug. 13 Room B Paper No: 1195
Flexible Job Shop Scheduling Problem by Chemical-reaction Optimization Algorithm
Junqing Li, Yuanzhen Li, Huaqing Yang, Kaizhou Gao, Yuting Wang and Tao Sun
School of Computer, Liaocheng University, Liaocheng, 252059
In this paper, we propose a novel discrete chemical-reaction optimization (DCRO) algorithm
for solving the flexible job shop scheduling problem with three objectives. The molecule is used
to represent a solution. The four elementary reactions, i.e., the on-wall ineffective collision, the
decomposition, the inter-molecular ineffective collision, and the synthesis, are used as the
68
operators for the hybrid algorithm. In the hybrid algorithm, the crossover operator is embedded
to learn information among molecules. To increase the ability to escape from the local optima,
the buffer is used as the energy center to share kinetic energy among molecules. Experimental
results on the well-known benchmarks show the efficiency and effectiveness of the proposed
algorithm. 17:30– 17:50 Afternoon Aug. 13 Room B Paper No: 1259
Ordinal Optimization-based Multi-Energy System Scheduling for Building Energy Saving
Zhong-Hua Su1, Qing-Shan Jia1,Chen Song2,
1 Center for Intelligent and Networked Systems, Department of Automation, Tsinghua
University, Beijing 100084, China,
2 Ubiquitous Energy Research Center
Buildings contribute a significant part in the energy consumption and CO2 emission in many
countries. Building energy saving has thus become a hot research topic recently. The technology
advances in power co-generation, on-site generation, and storage devices bring us the
opportunity to reduce the cost and CO2 emission while meeting the demand in buildings. A
fundamental difficulty to schedule this multi-energy system, besides other difficulties, is the
discrete and large search space. In this paper, the multi-energy scheduling problem is modeled
as a nonlinear programming problem with integer variables. A method is developed to solve this
problem in two steps, which uses ordinal optimization to address the discrete and large search
space and uses linear programming to solve the remaining sub-problems. The performance of
this method is theoretically quantified, and compared with enumeration and a
priority-and-rule-based scheduling policy. Numerical results show that our method provides a
good tradeoff between the solution quality and the computational time comparing with the other
two methods. We hope this work brings more insight on multi-energy scheduling problem in
general.
Room C Sino-Korean Intelligent Information Processing and Automation
Workshop III Chair Hee-Jun Kang, Hong-Hee Lee 13:40 – 13:53 Afternoon Aug. 13 Room C Paper No: 1069
Memristors by Quantum Mechanics
Thomas Prevenslik
QED Radiations, Disccovery Bay, Hong Kong, China
Memristor behavior is explained with a physical model based on quantum mechanics that
claims charge is naturally created anytime energy is absorbed at the nanoscale. Quantum
mechanics requires specific heat to vanish at the nanoscale, and therefore the electrical resistive
heating in the memristor cannot be conserved by an increase in temperature. Conservation
proceeds by frequency up-conversion of the absorbed energy to produce photons that in
submicron thin films have energy beyond the ultraviolet. By the photoelectric effect, the
photons create excitons inside the memristor that decrease resistance only to be recovered later
69
in the same cycle as the electrons and holes of the excitons are attracted to and destroyed by the
polarity of the voltage terminals. Observed memristor behavior is therefore the consequence of
excitons being created and destroyed every cycle. 13:53 – 14:06 Afternoon Aug. 13 Room C Paper No: 1861
A Robust Fault Detection and Isolation Scheme for Robot Manipulators based on Neural
Networks
Mien Van1, Hee-Jun Kang2 and Young-Shick Ro2
1 Graduate School of Electrical Engineering, University of Ulsan, 680-749, Ulsan, South
Korea
2 School of Electrical Engineering, University of Ulsan, 680-749, Ulsan, South Korea
This paper investigates an algorithm to the robust fault detection and isolation(FDI) in robot
manipulators using Neural Networks(NNs). Two Neural Networks are utilized: the first NN
(NN1) is employed to reproduce the robot’s dynamic behavior, while the second NN (NN2) is
used to achieve the online approximation for fault detection and isolation. This approach
focused on detecting changes in the robot dynamics due to faults. An online monitoring is used
not only to detect faults but also to provide estimates of the fault characteristics. A computer
simulation example for a two link robot manipulator shows the effectiveness of the proposed
algorithm in the fault detection and isolation design process. 14:06 – 14:19 Afternoon Aug. 13 Room C Paper No: 1306
A kernel function based estimation algorithm for multi-layer soil structure
Min-Jae Kang1, Chang-Jin Boo2 and Ho-Chan Kim2
1 Department of Electronic Engineering, Jeju National University, Korea
2 Department of Electrical Engineering, Jeju National University, Korea
This paper presents an analytic method based estimation scheme to extract soil parameters from
the kernel function of integral equation of apparent resistivity. A fast inversion method has been
developed for multi-layer earth structure based on the properties of kernel function. The
performance of the proposed method has been verified by carrying out a numerical example. 14:19 – 14:32 Afternoon Aug. 13 Room C Paper No: 1917
Automatic Context Analysis for Image Classification and Retrieval
Andrey Vavilin1, Kang-Hyun Jo1, Moon-Ho Jeong2, Jong-Eun Ha3 and Dong-Joong
Kang4
1Electrical Engineering Department, University of Ulsan {andy,
jkh2011}@islab.ulsan.ac.kr
2Kwangwoon University, [email protected]
3Seoul National University of Science, [email protected]
4Pusan National University, [email protected]
This paper describes a method for image classification and retrieval for natural and urban
scenes. The proposed algorithm is based on hierarchical image contents analysis. First image is
classified as urban or natural according to color and edge distribution properties. Additionally
scene is classified according to its conditions: illumination, weather, season and daytime based
on contrast, saturation and color properties of the image. Then image content is analyzed in
order to detect specific object classes: buildings, cars, trees, sky, road etc. To do so, image
recursively divided into rectangular blocks. For each block probabilities of membership in the
specific class is computed. This probability computed as a distance in a feature space defined by
70
optimal feature subset selected on the training step. Blocks which can not be assigned to any
class using computed features are separated into 4 sub-blocks which analyzed recursively.
Process stopped then all blocks are classified or size of block is smaller then predefined value.
Training process is used to select optimal feature subset for object classification. Training set
contains images with manually labeled objects of different classes. Each image additionally
tagged with scene parameters (illumination, weather etc). 14:32 – 14:45 Afternoon Aug. 13 Room C Paper No: 1772
A Discrete Artificial Bee Colony Algorithm for TSP Problem
Li Li1, Yurong Cheng1,Lijing Tan2, and Ben Niu1
1College of Management, Shenzhen University, Shenzhen 518060, China
2Management School, Jinan University, Guangzhou 510632, China
In this paper, a new discrete artificial bee colony algorithm is used to solve the symmetric
traveling salesman problem (TSP). The concept of Swap Operator has been introduced to the
original artificial bee colony (ABC) algorithm which can help the bees to generate a better
candidate tour by greedy selection. By taken six typical TSP instances as examples, the
proposed algorithm is compared with particle swarm optimization (PSO) algorithm to validate
its performance. In the experimental study, we also analysis the important parameters in the
artificial bee colony algorithm and their influence have been verified. 14:45 – 14:58 Afternoon Aug. 13 Room C Paper No: 1854
An Effective ECG Arrhythmia Classification Algorithm
Hsieh-Wei Chen1, Kuan-Rong Lee2, Hsun-Hui Huang1, Yaw-Huang Kuo1
Ya-Ting C. Yang2 and Yu-Liang Hsu1
1Department of Electrical Engineering
2Institute of Education & Center for Teacher Education, National Cheng Kung
University, Tainan 701, Taiwan, R.O.C.
This paper presents an effective electrocardiogram (ECG) arrhythmia classification scheme
consisting of a feature reduction method combining principal component analysis (PCA) with
linear discriminant analysis (LDA), and a probabilistic neural network (PNN) classifier to
discriminate eight different types of arrhythmia from ECG beats. Each ECG beat sample
composed of 200 sampling points at a 360 Hz sampling rate around an R peak is extracted from
ECG signals. The feature reduction method is employed to find important features from ECG
beats, and to improve the classification accuracy of the classifier. With the features, the PNN is
then trained to serve as classifier for discriminating eight different types of ECG beats. The
average classification accuracy of the proposed scheme is 99.71%. Our experimental results
have successfully validated the integration of the PNN classifier with the proposed feature
reduction method can achieve satisfactory classification accuracy. 14:58– 15:11 Afternoon Aug. 13 Room C Paper No: 1902
A Wearable Physical Activity Sensor System: Its Classification Algorithm and
Performance Comparison of Different Sensor Placements
Jeen-Shing Wang1, Fang-Chen Chuang1 and Ya-Ting C. Yang2
1Department of Electrical Engineering,
2Institute of Education & Center for Teacher Education, National Cheng Kung
University, Tainan, Taiwan, R.O.C.
This paper presents a wearable physical activity sensor system and its activity classification
71
algorithm. In addition, we investigate possible combinations of different sensor placements, and
identify an optimal combination to achieve the best classification performance. The sensor
system consists of several sensor modules that can be synchronized to record the accelerations
of diverse motions/activities. In our experiment, three sensor modules are mounted on
participants’ hand wrists, waists, and ankles, respectively, to collect seven categories of activity
accelerations. The proposed classification algorithm consisting of acceleration acquisition,
signal preprocessing, feature generation, and feature reduction, is capable of translating
time-series acceleration signals into important time- and frequency-domain feature vectors. The
dimension of features is reduced by linear discriminate analysis (LDA), and then the reduced
features are sent to a k-nearest neighbor (k-NN) classifier for classification. Our experimental
results have successfully validated the effectiveness of the proposed classification algorithm.
The best classification accuracy is 96.98% when the sensor modules are placed on hand and
ankle simultaneously. 15:11 – 15:25 Afternoon Aug. 13 Room C Paper No: 1911
Multi-objective Optimization Using BFO Algorithm
Ben Niu1,2,4, Hong Wang2, Lijing Tan3, Jun Xu4
1Hefei Institute of Intelligent Machines, Chinese Academy of Sciences, Hefei 230031,
China
2College of Management, Shenzhen University, Shenzhen 518060, China
3Management School, Jinan University, Guangzhou 510632, China
4e-Business Technology Institute, The University of Hongkong, Hongkong, China
This paper describes a novel bacterial foraging optimization (BFO) approach to multi-objective
optimization, called Multi-objective Bacterial Foraging Optimization (MBFO). The search for
Pareto optimal set of multi-objective optimization problems is implemented. Compared with the
proposed algorithm MOPSO and NSGAII, simulation results (measured by Diversity and
Generational Distance metric) on test problems show that the proposed MBFO is able to find a
much better spread of solutions and faster convergence to the true Pareto-optimal front. It
suggests that the proposed MBFO is very promising in dealing with multi-objective
optimization problems. 15:25– 15:40 Afternoon Aug. 13 Room C Paper No: 1922
A PACE Sensor System with Machine Learning-based Energy Expenditure Regression
Algorithm
Jeen-Shing Wang1, Che-Wei Lin1, Ya-Ting C. Yang2, Tzu-Ping Kao1, Wei-Hsin Wang1,
and Yen-Shiun Chen1
1Department of Electrical Engineering, 2Institute of Education & Center for Teacher
Education, National Cheng Kung University Tainan 701, Taiwan, R.O.C.
This paper presents a portable-accelerometer and electrocardiogram (PACE) sensor system and
a machine learning-based energy expenditure regression algorithm. The PACE sensor system
includes motion sensors and an electrocardiogram sensor, a MCU module (microcontroller), a
wireless communication module (a RF transceiver and a Bluetooth® module), and a storage
module (flash memory). A machine learning-based energy expenditure regression algorithm
consisting of the procedures of data collection, data preprocessing, feature selection, and
construction of energy expenditure regression model has been developed in this study. The
sequential forward search and the sequential backward search were employed as the feature
72
selection strategies, and a generalized regression neural network were employed as the energy
expenditure regression models in this study. Our experimental results exhibited that the
proposed machine learning-based energy expenditure regression algorithm can achieve
satisfactory energy expenditure estimation by combing appropriate feature selection technique
with machine learning-based regression models. 15:40 – 15:50 Room C Coffee Break
Sino-Korean Intelligent Information Processing and Automation
Workshop IV Chair Hong-Hee Lee, Myung Jae Yi 15:50 – 16:07 Afternoon Aug. 13 Room C Paper No: 1592
Implementation of High-Performance Sound Synthesis Engine for Plucked-String
Instruments
Myeongsu Kang1, Jiwon Choi1, Yongmin Kim1, Cheol-Hong Kim2 and Jong-Myon Kim1
1 School of Electrical Engineering, University of Ulsan, Ulsan, South Korea
2 School of Electronics and Computer Engineering, Chonnam National University,
Kwangju, South Korea
Among all types of production environment, identical parallel machines are frequently used to
increase the manufacturing capacity of the drilling operation in Taiwan printed circuit board
(PCB) industries. So when a manager plans the production scheduling, multiple but
conflicting objectives are often considered. Unlike the single objective problem, the
multiple-objective version no longer looks for an individual optimal solution, but a Pareto front
consisting of a set of non-dominated solutions. The manager then can select one of the
alternatives from the set. For this matter, our research aims at applying a variable
neighborhood search (VNS) algorithm in the identical parallel machine scheduling problem
(IPMSP) with two conflicting objectives: makespan and total tardiness. In VNS, two
neighborhoods are defined – insert a job to a different position or swap two jobs in the
sequence. To save the computational expense, one of the neighborhoods is randomly selected
for the target solution which is also arbitrarily chosen from the current Pareto front. The
proposed VNS algorithm is tested on a set of real data collected from a leading PCB factory in
Taiwan and its performance is compared with well-known methods in the literature. The
computational results show that VNS outperforms all competing algorithms – SPGA, MOGA,
NSGA-II, SPEA-II, and MACO in terms of solution quality and computational time. 16:07 – 16:24 Afternoon Aug. 13 Room C Paper No:
An Online Self Gain Tuning Computed Torque Controller for a Five-bar Manipulator
Tien Dung Le1, Hee-Jun Kang2* , Young-Soo Suh2
1 Graduate School of Electrical Engineering, University of Ulsan, 680-749, Ulsan, South
Korea
2 School of Electrical Engineering, University of Ulsan, 680-749, Ulsan, South Korea
73
1535 Parallel manipulators have advantages like high accuracy, high stiffness, high payload
capability, low moving inertia, and so on. This paper presents the problems of control the
five-bar manipulators using computed torque control method. In order to improve the control
performance, an online self gain tuning method using neural networks is proposed for gain
tuning of computed torque controller. Simulation results show the effectiveness of the proposed
method in comparison with the traditional computed torque control method. 16:24 – 16:41 Afternoon Aug. 13 Room C Paper No: 1645
Towards an Efficient Discovery Services in OPC Unified Architecture
Mai Son and Myeong-Jae Yi
School of Computer Engineering and Information Technology
University of Ulsan, San-29, Moogu-2 Dong, Namgu, Ulsan 680-749, South Korea
The OPC Unified Architecture, simply abbreviated to OPC UA, is the new generation of
well-known and globally successful OPC standards. The OPC UA has totally thirteen parts
which comprise specifications to ensure high performance communication, independent
platform, and unified data model in enterprise systems. Discovery part, one of OPC UA parts,
specifies a set of discovery services by which OPC UA Client perform discovery process to
obtain information about OPC UA servers, including endpoint and security information.
Unfortunately, the key part is still being developed, which leads to the limitation in
development of OPC UA components. In this paper, we not only enrich the discovery concept
by using WS-Discovery and extend it for global discovery case but also give readers an
approach of how to implement discovery services in practice. The implementation will be
deployed on Microsoft WCF Framework. 16:41 – 16:58 Afternoon Aug. 13 Room C Paper No: 1236
Active and Passive Nearest Neighbor Algorithm: a Newly-Developed Supervised Classifier
KaiYan Feng1;2, JunHui Gao1, KaiRui Feng3, Lei Liu1;2, and YiXue Li1;2
1Shanghai Center for Bioinformatics Technology 100, Qinzhou Road, Shanghai, China
2Key Laboratory of System Biology, Shanghai Institute for Biological Sciences, Chinese
Academy of Sciences 320 Yueyang Road, Shanghai, China
3Simcyp Limited, Blades Enterprise Centre, John Street, Sheffield S2 4SU, United
Kingdom
K nearest neighbor algorithm (k-NN) is an instance-based lazy classifier that does not need to
delineate the entire boundaries between classes. Thus some classification tasks that constantly
need a training procedure may favor k-NN if high efficiency is needed. However, k-NN is
prone to be affected by the underlying data distribution. In this paper, we define a new
neighborhood relationship, called passive nearest neighbors, which is deemed to be able to
counteract with the variation of data densities. Based on which we develop a new classifier
called active and passive nearest neighbor algorithm (APNNA). The classifier is evaluated by
10-fold cross-validation on 10 randomly chosen benchmark datasets. The experimental results
show that APNNA performs better than other classifiers on some datasets and worse on some
other datasets, indicating that APNNA is a good complement to the current state-of-the-art of
classification. 16:58 – 17:15 Afternoon Aug. 13
Globe Robust Stability Analysis for Interval Neutral Systems
Duyu Liu and Xin Gao
College of Electrical and Information Engineering, Southwest University for
74
Room C Paper No: 1247
Nationalities.
In this paper, the robust asymptotical stability is investigated for a class of interval neutral
systems. Based on Lyapunov stable theory, the delay-dependent criteria are derived to ensure
the global, robust, asymptotical stability of the addressed system. The criteria can be checked
easily by LMI control toolbox in Matlab. Two numeric examples are given to illustrate the
effectiveness and improvement over some existing results. 17:15 – 17:32 Afternoon Aug. 13 Room C Paper No: 1326
Semantic Pattern-Based User Interactive Question Answering: User Interface Design and
Evaluation
Tianyong Hao1, Wenyin Liu 2 and Chunshen Zhu1
1 Department of Chinese, Translation and Linguistics,City University of Hong Kong
2 Department of Computer Science,City University of Hong Kong
This paper presents two user interfaces for a pattern-based User Interactive Question Answering
system, which are designed to encourage users to ask questions by using semantic patterns. One
is a Guide-Based User Interface (GBUI), which can guide users with clear instructions.
However, it involves many steps and the operation may become tedious. The other is a
Recommendation-Based User Interface (RBUI), which recommends a few relevant patterns
containing automatically suggested details for each free-text question. However, the
recommended patterns may not always be satisfactory and sometimes the user’s revision is
needed. In comparing these two user interfaces, we propose a new Complexity Evaluation
Model (CEM) to evaluate the complexity on the basis of user log study and a realistic focused
user study. The results of the study user logs, which cover a test set of 1605 users and 488
semantic patterns, show that RBUI can improve the complexity of GBUI by 39.8% on average.
The improvement is also confirmed by the user study. It has thus become clear that RBUI can
improve the usability of the UIQA system in terms of helping the system accumulate high
quality pattern-based questions. 17:32 – 17:50 Afternoon Aug. 13 Room C Paper No: 1324
Studies on the Automatic Recognition of Modern Chinese Conjunction Usages Hongying Zan, Lijuan Zhou and Kunli Zhang
School of Information Engineering, Zhengzhou University, 450001, Zhengzhou, China
The conjunctions can connect words, sentences and even paragraphs. They have special
connection functions and their usages are complex and diverse. At present, the studies on
conjunctions are mostly human-oriented. These descriptions can not avoid such limitations as
subjectivity and illegibility, and are not easy to be applied directly to natural language
processing (NLP). This paper studies the automatic recognition of conjunction usages in the
background of NLP. It designs a rule-based method and several statistical methods for conjunction
usages recognition. Results are compared and analyzed and turns out that rule-based method and
statistical methods have advantages and disadvantages.