Consistent Multiple Graph Embedding for Multi-View Clustering

Graph-based multi-view clustering aiming to obtain a partition of data across multiple views, has received considerable attention in recent years. Although great efforts have been made for graph-based clustering, it is still challenging fuse characteristics from various views learn common representation clustering. In this paper, we propose novel Consistent Multiple Graph Embedding Clustering framework (CMGEC). Specifically, graph auto-encoder (M-GAE) designed flexibly encode the complementary information using multi-graph fusion encoder. To guide learned maintaining similarity neighboring each view, Multi-view Mutual Information Maximization module (MMIM) introduced. Furthermore, network (GFN) devised explore relationship among graphs different and provide consensus needed M-GAE. By jointly training these models, can be obtained, which encodes more depicts comprehensively. Experiments on three types datasets demonstrate CMGEC outperforms state-of-the-art methods.