Kernel-based learning and feature selection analysis for cancer diagnosis

详细信息    查看全文

• 作者：Seyyid Ahmed Medjahed^a ; ^{seyyid.ahmed@univ-usto.dz} ; ^{seyyidahmed.medjahed@univ-usto.dz} ; Tamazouzt Ait Saadi^b ; ^{tamazouzt.ait.saadi@univ-lehavre.fr} ; Abdelkader Benyettou^a ; ^{aek.benyettou@univ-usto.dz} ; Mohammed Ouali^c ; ^d ; ^{mohammed.ouali@usherbrooke.ca}
• 关键词：Classification ; Feature selection ; Kernel-based learning ; Support vector machines recursive feature elimination ; Binary dragon fly
• 刊名：Applied Soft Computing
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：51
• 期：Complete
• 页码：39-48
• 全文大小：1176 K
• 卷排序：51

文摘

A novel feature selection approach is proposed based on two steps. First step uses SVM-RFE to prefiltre the gene; we select 60% of relevant genes. Second step uses Binary Dragon Fly algorithm to optimal subset of genes. Objective function is the average of classification rate of three Kernel-based classifiers. The numerical results show the efficacy of the proposed approach.