用于数据挖掘的粗集产生多重知识库的方法

提出了一种用于数据挖掘的粗集产生多重知识库的方法．该方法基于粗集理论的简化和决策矩阵．首先从知识表示系统中消除冗余属性，产生简化表，形成简化的知识表示系统，提高学习的有效性和精确性．然后从简化表中推导出决策矩阵，通过决策矩阵获取最小决策规则，计算包括所有必需属性的简化集合，给出计算最小决策规则和计算多重简化的算法．最后由相应的简化对最小决策规则进行分组形成多重知识库，给出产生多重知识库的算法．

Generating Multiple Knowledge Bases Using Rough Set in Data Mining

A new approach to generate multiple knowledge using rough set in data mining is presen ted, which is based on the reduction and decision matrix of the rough set theory. It first eliminates the superfluous attributes from the knowledge representation system to produce a reduction table and form a simple knowledge representation system to improve the efficiency and accuracy of the learning. Next, the decision matrices are derived from the reduction table and the minimal decision rules are obtained through the decision matrices. A set of reduction which include all the indispensable attributes is computed. Algorithms which compute the minimal decision rules and multiple reduction are given. Finally, the minimal decision rules are grouped to the corresponding reduction to form multiple knowledge bases and the algorithm of generating multiple knowledge bases is given.