Mining Representative Subset Based on Fuzzy Clustering  

作　　者：ZHOU Hongfang FENG Boqin LU Lintao 

机构地区：[1]School of Computer Science and Engineering, Xi＇an University of Technology, Xi＇an 710048, Shaanxi, China [2]School of Electronics and Information Engineering, Xi＇an Jiaotong University, Xi＇an 710049, Shaanxi, China

出　　处：《Wuhan University Journal of Natural Sciences》2007年第5期799-803,共5页武汉大学学报（自然科学英文版）

基　　金：Supported by the National High Technology Research and Development Program of China (2001AA113182)

摘　　要：Two new concepts-fuzzy mutuality and average fuzzy entropy are presented. Then based on these concepts, a new algorithm-RSMA （representative subset mining algorithm） is proposed, which can abstract representative subset from massive data. To accelerate the speed of producing representative subset, an improved algorithm-ARSMA（accelerated representative subset mining algorithm） is advanced, which adopt combining putting forward with backward strategies. In this way, the performance of the algorithm is improved. Finally we make experiments on real datasets and evaluate the representative subset. The experiment shows that ARSMA algorithm is more excellent than RandomPick algorithm either on effectiveness or efficiency.Two new concepts-fuzzy mutuality and average fuzzy entropy are presented. Then based on these concepts, a new algorithm-RSMA （representative subset mining algorithm） is proposed, which can abstract representative subset from massive data. To accelerate the speed of producing representative subset, an improved algorithm-ARSMA（accelerated representative subset mining algorithm） is advanced, which adopt combining putting forward with backward strategies. In this way, the performance of the algorithm is improved. Finally we make experiments on real datasets and evaluate the representative subset. The experiment shows that ARSMA algorithm is more excellent than RandomPick algorithm either on effectiveness or efficiency.

关 键 词：representative subset fuzzy mutuality fuzzy entropy COVERAGE REDUNDANCY 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]