Academic Journal
Contrastive graph clustering with Structure-Robust learning and stable prototype guidance.
| Τίτλος: | Contrastive graph clustering with Structure-Robust learning and stable prototype guidance. |
|---|---|
| Συγγραφείς: | Ding Y; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China. Electronic address: 930507586@qq.com., Xu X; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China. Electronic address: xxf@ycit.cn., Xu S; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China; College of Information and Communication Engineering, Harbin Engineering University, Harbin, 150001, China. Electronic address: xusen@ycit.cn., Bian X; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China. Electronic address: xsbian@ycit.edu.cn., Yao S; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China. Electronic address: Shanliang.Yao@ycit.edu.cn., Guo N; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China. Electronic address: guonaixuan@ycit.edu.cn., Wang W; School of Information Engineering, Yancheng Institute of Technology, Yancheng, 224051, China; Jiangsu Provincial Engineering Technology Center for Multimodal Perception and Intelligent Control of Offshore Wind Power Systems, Yancheng, 224051, China. Electronic address: wangwei@ycit.edu.cn., Zhou T; College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin, 150001, China. Electronic address: zhoutian@hrbeu.edu.cn., Shen Y; Jiangsu Tingshan Technology Co., Ltd, Yancheng, 224051, Jiangsu, China. Electronic address: shone@tingshan.cn. |
| Πηγή: | Neural networks : the official journal of the International Neural Network Society [Neural Netw] 2026 Oct; Vol. 202, pp. 109056. Date of Electronic Publication: 2026 Apr 29. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: Pergamon Press Country of Publication: United States NLM ID: 8805018 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1879-2782 (Electronic) Linking ISSN: 08936080 NLM ISO Abbreviation: Neural Netw Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: New York : Pergamon Press, [c1988- |
| Ιατρικοί όροι (MeSH): | Algorithms* , Cluster Analysis*, Representation Machine Learning |
| Περίληψη: | Contrastive graph clustering is an effective paradigm for unsupervised node representation learning, yet it still faces two challenges in high-dimensional attributed graphs. First, globally shared feature weights ignore node-wise variations in feature importance, reducing inter-cluster separability. Second, target distributions derived solely from current predictions lack temporal consistency, causing assignment fluctuations and prototype drift. To address these issues, we propose a unified framework termed Contrastive Graph Clustering with Structure-Robust Learning and Stable Prototype Guidance (SR-SPG). SR-SPG performs node-adaptive feature gating by integrating three priors: the neighborhood consistency prior, Fisher-discriminative prior, and adaptive learning prior, thereby enhancing feature separability. It also constructs temporally smoothed self-training targets via EMA-updated prototypes, thereby improving training stability. Extensive experiments on six benchmark datasets demonstrate the effectiveness and robustness of SR-SPG. (Copyright © 2026 Elsevier Ltd. All rights reserved.) |
| Competing Interests: | Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. |
| Contributed Indexing: | Keywords: Attributed graph clustering; Contrastive graph clustering; Node-adaptive feature gating; Self-supervised learning; Stable prototype guidance |
| Entry Date(s): | Date Created: 20260506 Date Completed: 20260718 Latest Revision: 20260727 |
| Update Code: | 20260728 |
| DOI: | 10.1016/j.neunet.2026.109056 |
| PMID: | 42090868 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 1879-2782 |
|---|---|
| DOI: | 10.1016/j.neunet.2026.109056 |