.906 北京科技大学学报 2006年第9期 参考文献 [5]Jelonek J.Krawiee K,Slowinski R.Rough set reduction of [1]Pawlak Z.Rough set.Int J Comput Inf Sci.1982(1)341 attributes and their domains for neural networks.Comput In- tel,1995,11(2):339 [2]Krysikiewicz M.Rough set approach to incomplete informa- tion system.Inf Sci.1998.112:39 [6]苗夺谦.Rough Set理论中连续属性的离散化方法.自动化 学报，2001,27(3)：296 [3]赵军，王国胤。基于粗集理论的数据离散化方法，小型微 型计算机系统，2004,25(1)：60 [7]Nguyen HS.Skowron A.Boolean reasoning for feature ex- [4]鄂旭，高学东，武森，等.信息表中不完备数据的填补方 traction problems/10th International Symposium on Foun 法.北京科技大学学报，2005,27(3)：364 dations of Intelligent Systems.New York.1997:116 A method for preprocessing an incomplete information table EXu),GAO Xuedong),SHAO Liangshan,YE Baiqing 1)Department of Computer Science.Liaoning Institute of Technology.Jinzhou 121001,China 2)Management School.University of Science and Technology Beijing.Beijing 100083.China 3)Management School,Liaoning Technical University.Fuxin 123000.China ABSTRACI This paper studied the problems of filling up incomplete data,reducing redundant attributes and discretizing continuous attributes in preprocessing the incomplete information table with continuous at- tributes in a rough set.According to the concept of interval value and the consistency of condition attributes and decision attributes,a plus rule for interval values was defined to filling up the incomplete data.Depend- ing on the conception of classification,the discernible vector was defined and the discernible vector addition rule was used to delete redundant attributes.By use of the super-club data and entropy of the information table,the discretization of continuous attributes was implemented.The illustration and experimental results indicate that the method is effective. KEY WORDS incomplete information table:rough set:information entropy;attributes reduction;dis- cretization参 考 文 献 ［1］ Pawlak Z．Rough set．Int J Comput Inf Sci‚1982（1）：341 ［2］ Krysikiewicz M．Rough set approach to incomplete informa￾tion system．Inf Sci‚1998‚112：39 ［3］ 赵军‚王国胤．基于粗集理论的数据离散化方法．小型微 型计算机系统‚2004‚25（1）：60 ［4］ 鄂旭‚高学东‚武森‚等．信息表中不完备数据的填补方 法．北京科技大学学报‚2005‚27（3）：364 ［5］ Jelonek J‚Krawiec K‚Slowinski R．Rough set reduction of attributes and their domains for neural networks．Comput In￾tell‚1995‚11（2）：339 ［6］ 苗夺谦．Rough Set 理论中连续属性的离散化方法．自动化 学报‚2001‚27（3）：296 ［7］ Nguyen H S‚Skowron A．Boolean reasoning for feature ex￾traction problems ∥ 10th International Symposium on Foun￾dations of Intelligent Systems．New York‚1997：116 A method for preprocessing an incomplete information table E Xu 1‚2）‚GAO Xuedong 1）‚SHAO L iangshan 3）‚Y E Baiqing 3） 1） Department of Computer Science‚Liaoning Institute of Technology‚Jinzhou121001‚China 2） Management School‚University of Science and Technology Beijing‚Beijing100083‚China 3） Management School‚Liaoning Technical University‚Fuxin123000‚China ABSTRACT This paper studied the problems of filling up incomplete data‚reducing redundant attributes and discretizing continuous attributes in preprocessing the incomplete information table with continuous at￾tributes in a rough set．According to the concept of interval value and the consistency of condition attributes and decision attributes‚a plus rule for interval values was defined to filling up the incomplete data．Depend￾ing on the conception of classification‚the discernible vector was defined and the discernible vector addition rule was used to delete redundant attributes．By use of the super-club data and entropy of the information table‚the discretization of continuous attributes was implemented．The illustration and experimental results indicate that the method is effective． KEY WORDS incomplete information table；rough set；information entropy；attributes reduction；dis￾cretization ·906· 北 京 科 技 大 学 学 报 2006年第9期