Analyzing Correlation Coefficients of Objective Rule Evaluation Indices on Classification Rules
In data mining post-processing,which is one of important procedures in a data mining process,at least 39 metrics have been proposed to find out valuable knowledge.However,their functional properties have never been clearly articulated under the same condition.Therefore,we carried out a correlation analysis of functional properties between each objective rule evaluation indices on classification rule sets using correlation coefficients between each index.In this analysis,we calculated average values of each index using bootstrap method on 34 classification rule sets learned based on information gain ratio.Then,we found the following relationships based on correlation coefficient values:similar pairs,discrepant pairs,and independent indices.With regarding to this result,we discuss about relative functional relationships between each group of objective indices.
Data Mining Post-processing Rule Evaluation Index Cor-relation Analysis
Hidenao Abe Shusaku Tsumoto
Shimane University 89-1 Enya-cho Izumo Shimane,6938501,Japan
国际会议
成都
英文
467-474
2008-05-17（万方平台首次上网日期，不代表论文的发表时间）