Modified great deluge for attribute reduction in rough set theory

Attribute reduction can be defined as a process of selecting a minimal subset of attributes (based on a rough set theory as a mathematical tool) from an original set with least lose of information. In this work, a modified great deluge algorithm has been employed on attribute reduction problems, where the search space is divided into three regions. In each region, the water level is updated using a different scheme based on the quality of the current solution, instead of using a linear mechanism which is used in the original great deluge algorithm. The proposed approach is tested on 13 standard benchmark datasets and able to obtain promising results when compared to state-of-the-art approaches.