Sparse-Gradient-Regularized Multi-View Clustering with Joint ℓ2,log Sparsity and a Tensor γ* Rank Surrogate
Open Access
- 9 September 2026
- journal article
- research article
- Published by MDPI AG in Mathematics
- Vol. 14 (18) , 3268
Abstract
Multi-view clustering seeks a consensus partition from heterogeneous feature views while retaining view-specific information. We propose SGLog--MSC, a nonconvex multi-view subspace clustering framework that combines graph total variation, an penalty, and a tensor spectral penalty. These terms promote locally consistent self-representations, model sample-wise corruption, and capture shared low-rank structure across views, respectively. We derive an alternating augmented-Lagrangian algorithm with candidate-selection rules that return global minimizers for both nonconvex proximal subproblems. We also state explicit conditions under which accumulation points satisfy the KKT system and establish conditional whole-sequence convergence through the Kurdyka–Łojasiewicz framework. Hyperparameters are selected by a label-free protocol specified before the audited rerun; ground-truth labels are never used during selection. Across 20 recorded k-means++ initializations of each fixed embedding, the method attains NMI on Yale and on Scene-15. Relative to nine literature-reported baselines, these are the highest reported NMI values on the two datasets. On UCI digits, BBCSport, and ORL, the method reaches NMI , , and , respectively, ranking second only to HLR-M2VS in the descriptive cross-paper comparison. Component-wise ablation identifies the term as the largest and most consistent contributor, while the effects of the surrogate and sparse-gradient term depend on the dataset. Compared with the tensor nuclear norm, the surrogate improves performance on BBCSport, ORL, and UCI digits; sensitivity analysis shows that an interior also improves performance on Scene-15. Together, these results support combining robust error modeling with local and shared structural regularization while emphasizing the dataset dependence of individual components.
Keywords
Funding Information
- National Natural Science Foundation of China (12571401)
- Heilongjiang Provincial Natural Science Foundation Joint Guidance Project (JJ2024LH0820)
- Fundamental Research Funds for the Central Universities (30722024CFJ2403)
- Key Project of the Natural Science Foundation in Higher Education Institutions of Anhui Province (2023AH030081)
- Key Project of the Natural Science Foundation in Higher Education Institutions of Anhui Province (2022AH050393)
- Key Project of the Natural Science Foundation in Higher Education Institutions of Anhui Province (2025AHGXZK20015)
- University Synergy Innovation Program of Anhui Province (GXXT-2023-108)
This publication has 47 references indexed in Scilit:
- Nonconvex tensor multiview subspace clustering with bipartite graph regularizationPattern Recognition, 2025
- Tensorial multi-view subspace clustering based on logarithmic
n
-order penaltyPattern Recognition, 2025
- Partition-level fusion induced multi-view Subspace Clustering with Tensorial Geman RankNeural Networks, 2024
- Log-based sparse nonnegative matrix factorization for data representationKnowledge-Based Systems, 2022
- Hyper-Laplacian regularized multi-view subspace clustering with low-rank tensor constraintNeural Networks, 2020
- On Unifying Multi-view Self-Representations for Clustering by Tensor Multi-rank MinimizationInternational Journal of Computer Vision, 2018
- Low rank subspace clustering (LRSC)Pattern Recognition Letters, 2014
- Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward–backward splitting, and regularized Gauss–Seidel methodsMathematical Programming, 2011
- Factorization strategies for third-order tensorsLinear Algebra and its Applications, 2011
- Nonsmooth Analysis of Singular Values. Part I: TheorySet-Valued Analysis, 2005