不完备信息系统中知识获取算法  

Algorithms for Knowledge Acquisition in Incomplete Information System

作　　者：顾沈明[1] 吴伟志[2] 高济[3] 

机构地区：[1]浙江大学软件学院,杭州310027 [2]浙江海洋学院信息学院,浙江舟山316004 [3]浙江大学人工智能研究所,杭州310027

出　　处：《计算机科学》2005年第9期149-152,共4页Computer Science

基　　金：国家自然科学基金(60373078);浙江省教育厅科研计划项目(20040538)资助

摘　　要：粗糙集理论是一种新的处理模糊和不确定知识的软计算工具。应用粗糙集理论,可以将隐藏在系统的知识能够以决策规则的形式表达出来。根据粗糙集上下近似的概念,决策规则能够分成确定性规则和可能性规则两种。本文将介绍从不完备信息系统中知识获取的算法,通过这些算法能够从不完备决策表中生成一种确定性的规则和两种可能性的规则,同时也介绍了不完备决策表中描述约简的算法。Rough Set theory is emerging as a powerful tool for reasoning about data. Using Rough Set theory, knowledge hidden in Incomplete Information System may be unraveled and expressed in the form of decision rules. According to the lower and upper approximations, decision rules can be divided into certain and possible rules. Algorithms for knowledge acquisition in incomplete information systems are proposed. As the result, one type of “certain” and two types of “possible” decision rules are generated from incomplete decision tables. Algorithms for reduction of descriptors in such tables are also discussed.

关 键 词：不完备信息系统 粗糙集算法 决策规则 确定性 决策表 软计算工具 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]