决策风险最小化属性约简-计算机科学与探索.pdfVIP

ISSN 1673-9418 CODEN JKYTA8 E-mail: fcst@ Journal of Frontiers of Computer Science and Technology 1673-9418/2011/05(02)-0155-06 Tel: +86-10 DOI: 10.3778/j.issn.1673-9418.2011.02.005 * + 贾修一 , 商 琳, 陈家骏 南京大学 计算机软件新技术国家重点实验室, 南京 210093 * Attribute Reduction Based on Minimum Decision Cost + JIA Xiuyi , SHANG Lin, CHEN Jiajun State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210093, China + Corresponding author: E-mail: jiaxy@ JIA Xiuyi, SHANG Lin, CHEN Jiajun. Attribute reduction based on minimum decision cost. Journal of Fron- tiers of Computer Science and Technology, 2011, 5(2): 155-160. Abstract: Attribute reductions in decision-theoretic rough set model are generally defined based on keeping the positive region unchanged or the non-negative region unchanged. Because the positive region and non-negative re- gion are preferred to the negative region by people in these definitions, it is difficult to evaluate and interpret this kind of subjective reductions. In decision-theoretic rough set model, all regions are determined according to the Bayesian decision procedure, and the minimum cost is considered as the principle in the decision procedure. Based on the minimum cost of decisions for the decision table, an attribute reduction is proposed. The cost of decisions which are decided based on the reduction is minimal. The definition is easier to understand and interpret than those region-based reductions. Key words: decision-theoretic rough set model; attribute reduction; minimum cost 决策粗糙集模型下目前定义的属性约简都要求约简前后正区域保持不变或者非负区域不变等,