基于闵科夫斯基距离的特征权重自调节软子空间聚类算法  

Minkowski distance based soft subspace clustering with feature weight self-adjustment mechanism

作　　者：支晓斌[1] 许朝晖[2] Zhi Xiaobin;Xu Zhaohui(School of Science,X i5 an University of Posts & Telecommunications, Xi'an710121 , China;School of Telecommunication & Information Engineering, X i5 an University of Posts & Telecommunications,Xi'an 710121,China)

机构地区：[1]西安邮电大学理学院,西安710121 [2]西安邮电大学通信与信息工程学院,西安710121

出　　处：《计算机应用研究》2016年第9期2688-2692,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61102095);陕西省自然科学基金资助项目(2014JM8307);陕西省科技计划项目(2014KJXX-72);陕西省教育厅专项科研计划基金资助项目(14JK1661)

摘　　要：针对特征权重自调节软子空间聚类(soft subspace clustering with feature weight self-adjustment mechanism,SC-FWSA)算法使用欧氏距离,存在对数据适应性较差的问题,将SC-FWSA算法中的欧氏距离拓展为闵科夫斯基距离(Minkowski distance),提出一种基于闵科夫斯基距离的特征权重自调节软子空间聚类(Minkowski distance based soft subspace clustering with feature weight self-adjustment mechanism,MSC-FWSA)算法,MSC-FWSA算法有效提高了SC-FWSA算法对数据的适应性。若干真实数据集上的对比性实验验证了MSC-FWSA算法的有效性。Seeing that SC-FWSA clustering algorithm used Euclidean distance and had a poor adaptability to data,this paper extended Euclidean distance in SC-FWSA clustering algorithm to Minkowski distance, and proposed a Minkowski distance based soft subspace clustering with feature weight self-adjustment mechanism( MSC-FWSA) . MSC-FWSA improves the adaptability of the SC-FWSA algorithm to data. Comparative experiments presented on some real data sets show the effectiveness of the MSC-FWSA algorithm.

关 键 词：聚类 特征加权 软子空间聚类 闵科夫斯基距离 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP301.6[自动化与计算机技术—计算机科学与技术]