Adaptive Weighted Multi-View Clustering

TL;DR

This paper introduces WM-NMF, a novel multi-view clustering method that adaptively learns view-specific and observation-specific weights to improve clustering accuracy and robustness to noise.

Contribution

The paper proposes a weighted multi-view NMF algorithm that automatically learns view and observation weights, addressing limitations of equal weighting and empirical tuning.

Findings

Outperforms existing methods in clustering accuracy.

Effectively handles noisy multi-view data.

Improves view relevance assessment through learned weights.

Abstract

Learning multi-view data is an emerging problem in machine learning research, and nonnegative matrix factorization (NMF) is a popular dimensionality-reduction method for integrating information from multiple views. These views often provide not only consensus but also complementary information. However, most multi-view NMF algorithms assign equal weight to each view or tune the weight via line search empirically, which can be infeasible without any prior knowledge of the views or computationally expensive. In this paper, we propose a weighted multi-view NMF (WM-NMF) algorithm. In particular, we aim to address the critical technical gap, which is to learn both view-specific weight and observation-specific reconstruction weight to quantify each view's information content. The introduced weighting scheme can alleviate unnecessary views' adverse effects and enlarge the positive effects of the important views by assigning smaller and larger weights, respectively. Experimental results confirm the effectiveness and advantages of the proposed algorithm in terms of achieving better clustering performance and dealing with the noisy data compared to the existing algorithms.