Presentation Title: Comparative Study of Discretization Methods in Data Mining

Committee:

• Dr. Jianhua Chen(chair)
• Dr.Rahul Shah
• Dr.Jian Zhang

Date: December 4, 2008
Time: 11:00 AM
Location: Coates Hall,Room 256

Abstract:
The Data Mining and Knowledge Discovery is the process of analyzing large amounts of data, and finding potentially important patterns in that data. This project consists of implementation of Naïve Bayes, and Decision Tree data mining algorithms. In this project different discretization methods are implemented to handle the numeric attributes for the Naïve Bayes Classifier, and two entropy based methods, and a pre-discretization method is used for Decision Trees. These methods will be compared according to the accuracy. The User Interface provides the options to preview data before training, select attributes based on entropy, use a testing set to find the accuracy of the model, and predict the outputs for individual instances. The project report consists of the methodologies, design and implementation of system, and the analysis of the results obtained from different datasets. The application is developed using C#, and the Graphical User Interface is designed using Visual Studio.NET, and ASP.NET 2005.

All are invited.