[Colloq] [Hiring Talk] Tony Hu, Tues 11/13 1130am 149 CN

[Rachel Bates]

College of Computer Science Colloquium

presents

Tony Xiaohua Hu
DMW Software, San Mateo, CA

who will speak on:
Data Mining Based on Database Set Operations and Rough Set Theory

When: Tuesday, November 13th, 11:30am
Where:  149 Cullinane Hall

Abstract:

In this talk I will discuss a new data mining method based on
database set operations and rough set theory. Our method reduces data
horizontally and vertically before data mining algorithms are applied
to discover various knowledge rules.   I will first give an
introduction of rough set theory, and then show how to redefine the
key concepts (reduct, core attributes, superficial attributes) of
rough set theory in the context of database set operations in order
to improve the computational efficiency for data mining applications.
Next I will discuss how to combine rough set method with database set
operations for discretization, generalization and rough set feature
(horizontal and vertical data reduction.). Then I will present
various data mining algorithms implemented in the prototype system
DBROUGH-II, show some results on a real data set, and a new approach
to construct a good ensemble of classifiers to improve the learning
accuracy.  In the end I will talk about my future research directions
briefly and conclude my presentation with Q&A.

Host:  Bob Futrelle