Multiclass Cancer Classification Method Based on Biomimetic Pattern Recognition
The accurate cancer classification is significant in clinical diagnostics. Recently, different classification methods based on DNA microarray data have been proposed to classify cancers. However, as to the multiclass classification problem with small sample size, it has no perfect method yet. An automatic classification method of cancers based on biomimetic pattern recognition has been referred to this paper, which also has been used to study the automatic classification for multiclass cancers. To the public GCM data set, the average correct classification rate reaches 78％ under the condition that the correct rejection rate is 79.7％.
Qian Su Ku Wang Yaoguang Wei Dong An Shoujue Wang Ming Sun
College of Information and Electrical Engineering China Agriculture University Beijing,100083 P.R Ch College of Information and Electrical Engineering China Agriculture University Beijing, 100083 P.R C Institute of Semiconductors Chinese Academy of Science Beijing,100083 P.R China
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）