IMPROVED ACCURACY AND ROUGHNESS MEASURES FOR ROUGH SETS  

作　　者：ZhouYuming XuBaowen 

机构地区：[1]DepartmentofcomputerScience＆Engineering,SoutheastUniversity,Nanjing210096 [2]StateKeyLaboratoryofSoftwareEngineering,WuhanUniversity,Wuhan430072

出　　处：《Journal of Electronics(China)》2002年第3期320-323,共4页电子科学学刊（英文版）

基　　金：National Natural Science Foundation of China(60073012);Natural Sceience Foundation of Jiangsu, China(BK2001004);Visiting Scholar Foundation of Key Lab in Wuhan University

摘　　要：Accuracy and roughness, proposed by Pawlak(1982), might draw a conclusion inconsistent with our intuition in some cases. This letter analyzes the limitations in these measures and proposes improved accuracy and roughness measures based on information theory.Accuracy and roughness, proposed by Pawlak(1982), might draw a conclusion inconsistent with our intuition in some cases. This letter analyzes the limitations in these measures and proposes improved accuracy and roughness measures based on information theory.

关 键 词：Rough sets Equivalence relations UNCERTAINTY Information theory Excess entropy 

分 类 号：O159[理学—数学]