WPI Worcester Polytechnic Institute

Computer Science Department
------------------------------------------

CS 525D KNOWLEDGE DISCOVERY AND DATA MINING  
PROJECT 4 - Inductive Logic Programming. Spring 2004

PROF. CAROLINA RUIZ 

Due Date: April 6th 2004 at 3:00 pm  
------------------------------------------

PROJECT DESCRIPTION


PROJECT DESCRIPTION

For each of the 3 datasets from Project 1, select one of the dataset attributes as the target attribute for classification, and construct the most accurate set of classification rules you can for predicting that target attribute.

For the construction of classification rules we will use both the PRISM algorithm (implemented in the Weka system) and the FOIL algorithm.


PROJECT ASSIGNMENT

  1. Construct, using the PRISM algorithm implemented in the Weka system, the most accurate hypothesis (i.e. set of rules) you can to predict the target attribute of each of your datasets.

  2. Construct, using a FOIL-like algorithm, the most accurate hypothesis (i.e. set of rules) you can to predict the target attribute of each of your datasets. The following are guidelines to use a FOIL-like algorithm to mine your patterns:

  3. Compare the set of rules produced by the two algoritms (i.e., Prism and Foil).

REPORT AND DUE DATE