Towards Learnable Anchor for Deep Multi-View Clustering

2025-03-16

AI Summary

To address the limitations of fixed, suboptimal anchors misaligned with true cluster structures and high time complexity (O(n²)) in deep multi-view clustering, this paper proposes a learnable anchor mechanism that jointly optimizes anchors and clustering structure in linear time O(n). We introduce the first differentiable anchor learning paradigm: (i) Gaussian-inspired stochastic perturbation to enhance robustness; (ii) an integrated framework combining anchor graph convolution with cross-view mutual information maximization; and (iii) multi-view embedding alignment. Anchors are updated end-to-end in a fully adaptive manner, eliminating reliance on handcrafted initialization. Extensive experiments on multiple benchmark datasets demonstrate consistent superiority over state-of-the-art methods, achieving an average 3.2% improvement in clustering accuracy while maintaining computational efficiency and robustness.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Deep multi-view clustering incorporating graph learning has presented tremendous potential. Most methods encounter costly square time consumption w.r.t. data size. Theoretically, anchor-based graph learning can alleviate this limitation, but related deep models mainly rely on manual discretization approaches to select anchors, which indicates that 1) the anchors are fixed during model training and 2) they may deviate from the true cluster distribution. Consequently, the unreliable anchors may corrupt clustering results. In this paper, we propose the Deep Multi-view Anchor Clustering (DMAC) model that performs clustering in linear time. Concretely, the initial anchors are intervened by the positive-incentive noise sampled from Gaussian distribution, such that they can be optimized with a newly designed anchor learning loss, which promotes a clear relationship between samples and anchors. Afterwards, anchor graph convolution is devised to model the cluster structure formed by the anchors, and the mutual information maximization loss is built to provide cross-view clustering guidance. In this way, the learned anchors can better represent clusters. With the optimal anchors, the full sample graph is calculated to derive a discriminative embedding for clustering. Extensive experiments on several datasets demonstrate the superior performance and efficiency of DMAC compared to state-of-the-art competitors.

Problem

• High time complexity in deep multi-view clustering
• Fixed and unreliable anchors in existing methods
• Improving anchor selection and clustering accuracy

Innovation

• Linear time clustering via optimized anchor learning
• Anchor graph convolution for cluster structure modeling
• Mutual information maximization for cross-view guidance