A New Algorithm Based on Variable Precision Rough Set to Deal With Noise Data in Data Mining

摘要：

The noise data processing in data mining is a very important issue. In this paper a variable precision rough set model is proposed to deal with noise data. Based on variable precision rough set theory, we analyze the attribute significance from the point of dependence rate, and propose two kinds of heuristic algorithms; one computes core set firstly, then reduces the attributes; the other one abandons computation for a core and reduces attributes directly. Finally, we use an actual example which applies the second algorithm to verify it.

关键词： variable precision rough set attribute reduction attribute significance

作者: Yong Yang

作者单位: Department of Information Management, CNPC Guangzhou Training Center, China

会议类型: 国际会议

会议名称: 2010 International Conference on Information Technology and Industrial Engineering(2010年信息技术与工业工程国际学术会议 ITIE 2010)

会议地点: 武汉

会议语种:英文

页码: 1122-1125

在线出版日期: 2010-06-06（万方平台首次上网日期，不代表论文的发表时间）