A SURVEY ON VISUAL ANALYSIS OF MULTIVARIATE DATA

PhD Qualifying Examination


Title: "A SURVEY ON VISUAL ANALYSIS OF MULTIVARIATE DATA"

by

Mr. Yixian ZHENG


Abstract:

Multivariate data analysis is important for many applications. In the big 
data era, most datasets we collect are multivariate. They are often large, 
heterogeneous, and high dimensional, which pose special challenges for 
analysts. Visual analytics has emerged as a major approach for 
multivariate data analysis as it exploits both data mining methods and 
visualization techniques, and keeps humans in the loop of analysis.

In this report, we survey recent visual analytics techniques for 
multivariate data analysis. We first briefly introduce different data 
types of multivariate data, followed by a list of major analytical tasks. 
After that, some representative statistical and data mining methods (e.g., 
dimension reduction) and visualization techniques (e.g., parallel 
coordinates) are summarized. The major part of this survey focuses on 
visual analytics techniques for various analytical tasks (e.g., cluster 
analysis, regression, and classification) and the applications of these 
techniques in urban informatics and financial engineering.  Finally, we 
conclude with possible research topics and directions.


Date:			Wednesday, 15 October 2014

Time:                  	4:00pm - 6:00pm

Venue:                  Room 5563
                         Lifts 27/28

Committee Members:	Prof. Lionel Ni (Supervisor)
  			Dr. Huamin Qu (Supervisor)
 			Dr. Raymond Wong (Chairperson)
 			Dr. Qiong Luo
  			Prof. Chiew-Lan Tai


**** ALL are Welcome ****