自适应特征选择加权k子凸包分类  

Weighted k sub-convex-hull classifier based on adaptive feature selection

作　　者：牟廉明[1,2] MOU Lianming(College of Mathematics and Information Science,Neijiang Normal University,Neijiang 641100,Sichuan,China;Data Recovery Key Laboratory of Sichuan Province,Neijiang 641100,Sichuan,China)

机构地区：[1]内江师范学院数学与信息科学学院,四川内江641100 [2]数据恢复四川省重点实验室,四川内江641100

出　　处：《山东大学学报（工学版）》2018年第5期32-37,共6页Journal of Shandong University（Engineering Science）

基　　金：国家自然科学基金资助项目(10872085);四川省科技厅科技计划重点资助项目(2017JY0199);四川教育厅自然科学重点项目基金资助项目(13ZA0008);2015内江市科技支撑计划资助项目

摘　　要：针对问题维数的增加以及不同特征对分类的作用往往不一样,导致k子凸包分类性能降低等问题,设计自适应特征选择加权k子凸包分类方法。根据传统凸包距离存在的不足引入加权k子凸包距离,在测试样本的k邻域内引入距离度量学习技术和正则化技术进行自适应的特征选择,并将自适应特征选择无缝嵌入加权k子凸包优化模型中,这样就能为不同的测试样本在不同的类别中学习自适应特征空间,得到有效的加权k子凸包距离计算方法。试验结果表明,该方法不仅能够进行降维,而且具有明显的分类性能优势。Because of the increase of the dimension of the problem and the effect of different features on classifier,the performance of the k sub-convex-hull classifier was seriously reduced. An adaptive feature selection weighted k sub-convex-hull classifier was designed（ AWCH）. A weighted k sub-convex-hull classifier was designed according to the shortcomings of conventional convex hull distance. By applying the distance metric learning and regularization technique in the k neighborhood of the test sample,an adaptive feature selection method was designed and seamlessly integrated into the optimization model on the weighted k sub-convex-hull.Through these efforts,for different test samples,an adaptive feature space in different categories could be extracted,and a valid weighted k sub-convex-hull distance could be obtained. Experimental results showed that the AWCH not only reduced the dimension of the problem,but also was significantly superior to similar classifiers.

关 键 词：加权k子凸包 度量学习 正则化 特征选择 自适应 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]