AN EFFECTIVE ALGORITHM FOR COMPUTING REDUCTS IN DECISION TABLES

Attribute reduction is one important part researched in rough set theory. A reduct from a decision table minimal subset of the conditional attributes which provide same information for classification purposes as entire available attributes. The task high dimensional could be solved faster if reduct, instead original whole attributes, used. In this paper, we propose computing algorithm using attribute clustering. proposed works three main stages. first stage, irrelevant are eliminated. second stage relevant divided into appropriately selected number clusters by Partitioning Around Medoids (PAM) clustering method integrated with special metric space normalized variation information. third representative each cluster that most class-related. form approximate reduct. implemented and experimented. experimental results show capable small size accuracy, when used to group selected.