Genetic Algorithm for Feature Selection of MR Brain Images Using Wavelet Co-occurence

摘要：

The selection of features has a considerable impact on the success or failure of classification process. Feature selection refers to the procedure of selecting a subset of informative attributes to build models describing data. The main purpose of feature selection is to reduce the number of features used in classification while maintaining high classification accuracy. A large number of algorithms have been proposed for feature subset selection. Here we compare classical sequential methods with the genetic approach in terms of the number of features, classification accuracy and reduction rate. Genetic Algorithm (GA) achieves an acceptable classification accuracy with only five of the available 44 features. The optimal feature such as mean of contrast, mean of homogeneity, mean of sum average, mean of sum variance and range of autocorrelation provide best classification performance. Similar classification performance is obtained with SFFS and SFBS but with larger feature set.

关键词： Featureselection classification accuracy reduction rate Genetic Algorithm

作者: Ahmed Kharrat Mohamed Ben Messaoud Nacera Benamrane Mohamed Abid

作者单位: University of Sfax National Engineering School Computer & Embedded Systems Laboratory (CES) B.P 1173 University of Sfax National Engineering School Electronics and Information Technologies B.P 1173, Sf Department of Computer Science Faculty of Science Vision and Medical Imagery Laboratory U.S.T.O. B.P

会议类型: 国际会议

会议名称: 2010 International Conference on Signal and Information Processing(2010年IEEE信号与信息处理国际会议 ICSIP2010)

会议地点: 长沙

会议语种:英文

页码: 606-610

在线出版日期: 2010-12-14（万方平台首次上网日期，不代表论文的发表时间）