自动化学报2018，Vol.44Issue(12)：2160-2169,10.DOI:10.16383/j.aas.2018.c160636
基于多视图矩阵分解的聚类分析
Matrix Factorization for Multi-view Clustering
摘要
Abstract
In computer vision and pattern recognition fields, more and more data are represented by multiple views which describe different perspectives of the data. And multi-view learning methods are developed for ultilizing the information sufficiently. In this paper, we propose two novel clustering methods called MultiGNMF and MultiGSemiNMF, respectively, which are based on multiview learning with local graph regularization, where the innerview relatedness between data is taken into consideration. However, MultiGNMF is based on NMF, which only applies to non-negative matrix. To eliminate this limit, we propose MultiGSemiNMF based on SemiNMF, which is also applicable for negative matrix. The experimental results demonstrate the effectiveness of our proposed methods.关键词
多视图学习/聚类/矩阵分解/局部结构正则化Key words
Multi-view learning/clustering/matrix factorization/local graph regularization引用本文复制引用
张祎,孔祥维,王振帆,付海燕,李明..基于多视图矩阵分解的聚类分析[J].自动化学报,2018,44(12):2160-2169,10.基金项目
国家自然科学基金(61772111) （61772111）
国家自然科学基金创新研究群体科学基金(71421001)资助 （71421001）
