Multi-View Clustering With the Cooperation of Visible and Hidden Views

Multi-view data are becoming common in real-world applications and many multi-view clustering algorithms have thus been proposed. The existing usually focus on the cooperation of different visible views original space but neglect influence hidden information among these views, or they only consider views. therefore not efficient since available is fully exploited, particularly otherness consistency them. In practice, both very useful for effective analyses. this study, a Multi-View algorithm with Cooperation Visible Hidden i.e., MV-Co-VH, MV-Co-VH first projects multiple from spaces to by using non-negative matrix factorization obtain view data. Collaborative learning then implemented procedure based shared view. experimental results extensive experiments UCI datasets image show that performance proposed competitive even better than algorithms.