High-dimensional multi-view clustering methods
Alaeddine Zahir, Khalide Jbilou, Ahmed Ratnani
TL;DR
The paper addresses the challenge of clustering high-dimensional multi-view data by contrasting graph-based and subspace-based frameworks, with a focus on tensorized representations that capture high-order cross-view correlations. It surveys core methods across matrix- and tensor-based MVC, including adaptive graph learning, multi-view subspace learning, and TLMSC-style tensor approaches, and evaluates them on diverse benchmark datasets. Key findings show that tensorized MVC and adaptive graph learning often outperform traditional matrix-based methods in clustering accuracy and robustness, highlighting the importance of preserving inter-view structure and high-order interactions. The work underscores the practical impact of tensor representations for multi-view clustering, offering guidance on method selection and directions for future research in high-dimensional data fusion and clustering performance.
Abstract
Multi-view clustering has been widely used in recent years in comparison to single-view clustering, for clear reasons, as it offers more insights into the data, which has brought with it some challenges, such as how to combine these views or features. Most of recent work in this field focuses mainly on tensor representation instead of treating the data as simple matrices. This permits to deal with the high-order correlation between the data which the based matrix approach struggles to capture. Accordingly, we will examine and compare these approaches, particularly in two categories, namely graph-based clustering and subspace-based clustering. We will conduct and report experiments of the main clustering methods over a benchmark datasets.