Adaptive feature selection via a new version of support vector machine

摘要: This paper focuses on feature selection in classification. A new version of support vector machine (SVM) named p-norm ( $$p\in[0,1]$$ ) is proposed. Different from the standard SVM, $$(p\in[0,1])$$ normal decision plane used which leads to more sparse solution. Our model can not only select less features but also improve classification accuracy by adjusting parameter p. The numerical experiments results show that our SVM effective than some usual methods selection.