MULTIPLE VIEW CLUSTERING BASED ON IMPROVED SLACK EMBEDDING SPACE
In view of the lack of unified feature representation and defects of conservatism of traditional clustering methods,a multiple view clustering method based on improved slack embedding space is proposed.In a unified framework,a comprehensive potential embedding representation matrix,a global similarity matrix and an accurate index matrix were jointly learned.Furthermore,the constraint of global similarity matrix was slack,and an improved slack multiple view clustering embedding space was proposed,which made the proposed method have lower computational complexity and more correlation between data point pairs.The experimental results show that the proposed method can obtain more robust and more accurate clustering results.
国家科技期刊平台
NSTL
万方数据
维普
