Efficient Gene Selection with Rough Sets from Gene Expression Data

摘要：

The main challenge of gene selection from gene expression dataset is to reduce the redundant genes without affecting discernibility between objects.A pipelined approach combining feature ranking together with rough sets attribute reduction for gene selection is proposed.Feature ranking is used to narrow down the gene space as the first step,top ranked genes are selected;the minimal reduct is induced by rough sets to eliminate the redundant attributes.An exploration of this approach on Leukemia gene expression data is conducted and good results are obtained with no preprocessing to the data.The experiment results show that this approach is successful for selecting high discriminative genes for cancer classification task.

关键词： Gene selection Feature ranking Rough sets Attributes reduction

作者: Lijun Sun Duoqian Miao Hongyun Zhang

作者单位: Department of Computer Science and Technology,Tongji University,Shanghai,201804,P.R.Cilina Department of Computer Science and Technology,Tongji University,Shanghai,201804,P.R.China

会议类型: 国际会议

会议名称: The Third International Conference on Rough Sets and Knowledge Tevhnology(RSKT 2008)(第三届粗糙集与知识技术国际会议)

会议地点: 成都

会议语种:英文

页码: 164-171

在线出版日期: 2008-05-17（万方平台首次上网日期，不代表论文的发表时间）