Post on 17-Jan-2016
Intelligent Database Systems Lab
Presenter : Kung, Chien-Hao
Authors : Eghbal G. Mansoori
2011,IEEE
FRBC: A Fuzzy Rule-Based Clustering Algorithm
Intelligent Database Systems Lab
Outlines
MotivationObjectivesMethodologyExperimentsConclusionsComments
Intelligent Database Systems Lab
Motivation• Clustering response is a primitive
exploratory approach in data
analysis with little or no prior
knowledge.
• However, the main challenge for
most of clustering algorithms is their
necessity to know the number of
clusters for which to look.
Intelligent Database Systems Lab
Objectives• To overcome these restrictions, a novel fuzzy rule-
based clustering algorithm(FRBC) is proposed in this
paper.
• FRBC tries to automatically explore the potential
clusters in the data patterns.
Intelligent Database Systems Lab
Methodology-Fuzzy • Fuzzification
• Fuzzy Rule
• Fuzzy Inference Mechanism
• Defuzzifierion
Intelligent Database Systems Lab
MethodologyGenerate auxiliary data
Choose the best rule
Clustering
Regroup remained data
Intelligent Database Systems Lab
MethodologyGenerate auxiliary data
Choose the best rule
Clustering
Regroup remained data
Intelligent Database Systems Lab
MethodologyGenerate auxiliary data
Choose the best rule
Clustering
Regroup remained data
Intelligent Database Systems Lab
MethodologyGenerate auxiliary data
Choose the best rule
Clustering
Regroup remained data
Intelligent Database Systems Lab
MethodologyGenerate auxiliary data
Choose the best rule
Clustering
Regroup remained data
Intelligent Database Systems Lab
Experiment
Intelligent Database Systems Lab
Experiment
Intelligent Database Systems Lab
Experiment
Intelligent Database Systems Lab
Experiment
T=0.1 T=0.01
Intelligent Database Systems Lab
Experiment
Intelligent Database Systems Lab
Experiment
Intelligent Database Systems Lab
Conclusions• FRBC is a novel fuzzy rule-based clustering algorithm
to automatically explore the potential clusters.
• The clusters specified by fuzzy rules are human understandable with acceptable accuracy.
Intelligent Database Systems Lab
Comments• Advantages/drawbacks– This paper gives rich experiments for this method– But this method still has a parameter (threshold)
to control the number of clusters.• Applications– Clustering.