摘要
针对目前应用粗糙集相似关系理论与LEM2算法进行规则推理时获取规则较少以及规则简化程度不高的问题,提出了粗糙集非对称相似关系与近似集的计算方法,并对现有LEM2算法获取规则的过程进行了改进与补充,形成了一种新的基于非对称相似粗糙集的规则获取算法,以便从不完整信息中获取更多潜在规则。最后以实际算例对两种算法分别进行了测试并给出了结果对比分析,仿真结果表明新的规则获取算法在不改变原有信息集结构与内容的基础上具有更好的优化性能,能获得更好的优化结果。
To meet the challenge of quantity and simplicity in LEM2 algorithm, a new rule extraction algorithm based on asymmetrical similarity rough set is proposed. This new algorithm combines the use of asymmetrical similarity calculation and improved LEM2 algorithm to extract more rules from incomplete information. At last, an example is used to test and compare the results of different algorithms. The simulation result shows that the new rule extraction algorithm can achieve better solution without changing the structure and content of original information set compared to traditional LEM2 algorithm.
出处
《计算机仿真》
CSCD
2008年第10期110-113,共4页
Computer Simulation
基金
国家自然科学基金(70671066)
关键词
粗糙集
非对称相似关系
规则获取
Rough set
Asymmetrical similarity
Rule extraction
