Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

32
Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas

Transcript of Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Page 1: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Clustering and Research Works

Dr. Bernard Chen Ph.D.University of Central Arkansas

Page 2: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Outline

Clustering Data Science Future Works

Page 3: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Clustering Algorithms

There are two clustering algorithms we used in our approach:

K-means Clustering Fuzzy C-means Clustering

Page 4: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

K-means Clustering

Page 5: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

K-means Clustering

Page 6: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

K-means Clustering

Page 7: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

K-means Clustering

Page 8: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

K-means Clustering

Page 9: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 10: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 11: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 12: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 13: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 14: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 15: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Fuzzy C-means Clustering

Page 16: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Real World example

Page 17: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Outline

Clustering Data Science Future Works

Page 18: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Data Science wikipedia

Data science is the study of the generalizable extraction of knowledge from data.

It incorporates varying elements and builds on techniques and theories from many fields

wikipedia

Page 19: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Outline

Clustering Data Science Future Works

Page 20: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Data Science wikipedia

A practitioner of data science is called a data scientist.

Data scientists solve complex data problems through employing deep expertise in some scientific discipline.

It is generally expected that data scientists are able to work with various elements

Page 21: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Data Science wikipedia

Good data scientists are able to apply their skills to achieve a broad spectrum of end results.

the ability to find and interpret rich data sources, manage large amounts of data despite hardware,

software and bandwidth constraints, merge data sources together, ensure consistency of data-sets, create visualizations to aid in understanding data, build mathematical models using the data, present and communicate the data insights/findings

to specialists and scientists in their team and if required to a naive audience.

Page 22: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Outline

Clustering Data Science Future Works

Page 23: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Data Science in WINE

Once viewed as a luxury good, nowadays wine is increasingly enjoyed by a wider range of consumers.

Wine certification is generally assessed by physicochemical and sensory tests

Page 24: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

sensory tests Example: Chateau Latour 2010

http://www.wine.com/V6/Chateau-Latour-2010/wine/110508/detail.aspx

Page 25: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

sensory tests Among those expert reviews, we use “Wine

Spectator’s” version

"Unbelievably pure, with distilled cassis and plum fruit that cuts a very precise path, while embers of anise, violet and black cherry configure form a gorgeous backdrop. A bedrock of graphite structure should help this outlive other 2010s. Powerful, sleek and incredibly long. Not perfect, but very close. Best from 2020 through 2050."99 Points Wine Spectator

Page 26: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

sensory tests

Wine Spectator has the following advantages: Words are precise Well-known Famous for it’s Top 100 wine of the

year selection Well maintained database

Page 27: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Research Topic 1

Clustering on past 10 years Top 100 wine (1000 wines)

Challenges: Extract attributes from 1000 wine Clustering algorithm Analysis of the results

Page 28: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Research Topic 2 Multi-label (4 classes) Classification on

1000 wines, which composed of 250 wines for 4 category (95+, 90~94, 89~85, 85-)

Challenges: Classification algorithm 4 classes How to improve accuracy

Page 29: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Research Topic 3 Association Rules on region-specific

dataset (such as Napa) for attribute correlation and quality prediction.

Challenges: Association Rules algorithm Analysis of the results How to improve accuracy

Page 30: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Research Topic 4 Region Prediction (such as France vs

Italy), open for association rules or classification algorithms.

Challenges: More free-style (more suitable for

experienced researchers) Not only focus on accuracy, but also try to

tell the difference between the regions

Page 31: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Research Topic 5

Clustering + Classification for higher accuracy prediction.

Challenges: TWO type of algorithms More complex in understanding and

coding

Page 32: Clustering and Research Works Dr. Bernard Chen Ph.D. University of Central Arkansas.

Research Topic 6

Multi-label research: since we have multiple reviews available, how to use those information for data science research?

Challenges: Very flexible!!!