关键词: 粗糙集, 特征选择, 加权平均支持启发函数

Abstract: Feature selection is a valid technique for information-preserving data reduction in data analysis.Rough set theory provides a mathematical tool which can be used to discovery all possible feature subsets.This paper proposes a new rough set-based heuristic function called weighted average support heuristic.The main advantage is that it considers the overall quality of the set of potential rules.In another words，it considers the weighted average support of the rules for all decision classes.At last，the example proves this method is valid.

Key words: rough set, feature selection, weighted average support heuristic