Global Discretization Approach Based on Minimize Entropy in Rough Sets Classification

The discretization approach produces a qualitative attribute from a quantitative attribute. That has many advantages, such as data can be reduced and simplified. Using discrete attributes are usually more compact, shorter and more accurate than using continuous ones. The LEM2 (Learning from Examples Module, version 2) rule extracting algorithm is superior to other algorithms in rough set, that the deduce rule sets directly from data with symbolic and numerical attributes, but LEM2 requires pre-discretized data. Therefore, this study proposes a global discretization approach, which integrated minimize entropy principle approach and rough sets, to enhance accuracy rate and reduce number of rules for solving classification problems. The experimental results indicate that the proposed approach outperforms the listing approaches.