Review of multi-view subspace clustering

Abstract

Multi-view subspace clustering is a type of subspace clustering which combines with multi-view learning. It cannot only deal with the challenges of the big data and high dimensions, but also cluster multi-view data from multiple sources and observation angles according to some similarity measurement. Our paper is to introduce the theoretical basis and latest research progress of multi-view subspace clustering. First, we briefly describe the basic principles, research status and classification of subspace clustering, and also compare several kinds of subspace clustering algorithms; next, multi-view subspace clustering is described in detail, and the ideas of several clustering algorithms are analysed and summarised. After that, we elaborate the research status of multi-view subspace clustering. Finally, the application of multi-view subspace clustering in various fields' application is introduced. The purpose of this paper is for beginners to quickly know about the research status of multi-view subspace clustering and some ideas of typical algorithms.