TW-Co-k-means: Two-level weighted collaborative k-means for multi-view clustering

Multi-view clustering has attracted an increasing amount of attention in recent years due to its ability to analyze data from multiple sources or views. Despite significant success, there are still two challenging problems in multi-view clustering, namely, (i) how to satisfy the consistency across different views while preserving the diversity within each view, and (ii) how to weight the different views and the features in each view w.r.t. their importance to improve the clustering result. In this paper, to simultaneously tackle these two problems, we propose a novel multi-view clustering approach termed Two-level Weighted Collaborative k-means (TW-Co-k-means). A new objective function is designed for multi-view clustering, which exploits the distinctive information in each view while taking advantage of the complementariness and consistency across different views in a collaborative manner. The views and the features in each view are assigned with weights that reflect their importance. We introduce an iterative optimization method to optimize the objective function and thereby achieve the final clustering result. Experimental results on multiple real-world datasets demonstrate the effectiveness of our approach.