COMMERCIAL INDUSTRIAL MATHEMATICS ACTIVITIES
|October 1, 2014|
Seminar on Data Mining & Customer-Centric Marketing
Schedule - April 25
|1:30 - 1:40
||Bradd Hart, Deputy Director of
The Fields Institute
Welcome & Introduction
|1:40 - 2:00
||Urfan Sayed, Account Officer at
Update on Data & Services
|2:00 - 2:45
||Dr. Bobby Siu, President, Infoworth
Color Your World: Why Multicultural Marketing Makes Sense
|2:45 - 3:15
|3:15 - 4:00
||Prof. Jianhong Wu, York University
Projective Clustering of Data Sets in Large Dimensional Spaces
|4:00 - 4:30
||Dr. Susan Kular, Manifold Data Mining Inc.
Databases, Business Operations and CRM
Dr. Susan Kular
Dr. Bobby Siu
Canada has been going through some demographic shifts in the past few decades. The presentation examines some population and consumption statistics and notes the value of developing marketing strategies for multicultural groups.
He has a Ph.D. in sociology from Carleton University and has held academic appointments at McMaster University, Queen's University, Ryerson Polytechnic University, University of Western Ontario and York University. In addition, he has held various senior management posts with the Government of Ontario where he assumed responsibility for a wide range of projects involving multicultural communities.
Through his business activities and sitting on the boards of
various organizations, Dr. Siu has extensive linkages to the ethnocultural
communities. His views on diversity and marketing have been quoted
in Canadian Business and Strategy. He has extensive publications
on intercultural and diversity including Canadian Ethnic Studies,
Canadian HR Reporter, Marketing Magazine, Strategy, and the award-winning
Toronto Region Business and Market Guide.
Prof. JianHong Wu
Projective clustering seems to be the only realistic approach in pattern recognition and feature detection of data sets in high dimensional spaces, as points in such data sets are usually sparsely distributed in the full space. The essential difficulty arises as the set of dimensions and the set of data points have to be found SIMULTANEOUSLY so that the points are corollated with respect to the set of dimensions. We report some of our recent progress in the design and implementation of a neural network architecture, called Projective Adaptive Resonance Theory, for projective clustering. The algorithm and numerical simulations will be described, together with some detailed comparisons with existing algorithms for clustring of large data sets in high dimensional spaces.
For further information please call Manifold Data Mining Inc. at 416-760-8828 or email at firstname.lastname@example.org