Comparison of Information Theoretical Measures for Reduct Finding

The paper discusses the properties of an attribute selection criterion for building rough set reducts based on discernibility matrix and compares it with Shannon entropy and Gini index used for building decision trees. It has been shown theoretically and experimentally that entropy and Gini index tend to work better if the reduct is later used for prediction of previously unseen cases, and the criterion based on the discernibility matrix tends to work better for learning functional relationships where generalization is not an issue.