A Consensus Anchor-guided Hypergraph Framework for Incomplete Multi-view Clustering

Handling large-scale incomplete multi-view data poses a significant challenge in unsupervised representation learning. While anchor-based strategies have alleviated computational burdens, they typically rely on shallow bipartite graphs restricted to pairwise relations, failing to capture complex high-order correlations among samples. Furthermore, existing methods often treat observed and missing instances indiscriminately, ignoring the distributional shifts that lead to systematic bias in consensus anchor learning. To address these limitations, we propose a novel framework tailored for scalability and robustness, termed Hypergraph-Augmented Incomplete Multi-View Clustering (HA-IMVC). Unlike traditional approaches, HA-IMVC constructs a consensus anchor-guided hypergraph that explicitly models group-wise interactions, thereby preserving structural integrity even under high missing rates. Crucially, we incorporate a dual-adaptive reweighting mechanism that calibrates importance at both the view and sample levels. This strategy adaptively penalizes severely incomplete samples to mitigate bias while harmonizing inconsistent views. Extensive experiments on diverse benchmarks demonstrate that HA-IMVC achieves superior clustering accuracy and maintains high efficiency, even in scenarios characterized by severe data incompleteness.

This paper primarily addresses the challenge of enabling computers to automatically classify objects quickly and accurately even when faced with massive amounts of data and significant information gaps. Traditional methods are highly prone to breaking down when information is missing and are easily misled by erroneous data. In contrast, the innovative HA-IMVC algorithm introduced in this paper functions like a "group chat" that brings all relevant data together. Even if a single data point is missing critical information, stable connections are maintained through indirect relationships among the data points. Additionally, the algorithm intelligently "adapts to the situation," automatically reducing the influence of severely incomplete or noisy data. Experiments demonstrate that the algorithm not only maintains extremely high classification accuracy even under extreme conditions where up to 70% of data is missing, but also offers fast computation and low memory consumption. It addresses the pain point of traditional algorithms—where computers often crash due to excessive load when processing big data—making it highly suitable for practical applications involving massive datasets in the industrial sector.