Localized Sparse Incomplete Multi-view Clustering

Incomplete multi-view clustering, which aims to solve the clustering problem on incomplete data with partial view missing, has received more and attention in recent years. Although numerous methods have been developed, most of either cannot flexibly handle arbitrary missing views or do not consider negative factor information imbalance among views. Moreover, some fully explore local structure all To tackle these problems, this paper proposes a simple but effective method, named localized sparse (LSIMVC). Different from existing methods, LSIMVC intends learn structured consensus latent representation by optimizing regularized novel graph embedded matrix factorization model. Specifically, such model based factorization, l1 norm constraint is introduced obtain low-dimensional individual representations representation. embedding term works, our aggregates task learning into concise term. Furthermore, reduce learning, an adaptive weighted scheme LSIMVC. Finally, efficient optimization strategy given proposed Comprehensive experimental results performed six databases verify that performance superior state-of-the-art IMC approaches. The code available https://github.com/justsmart/LSIMVC.