Enhanced Federated Deep Multi-View Clustering Under Uncertainty Scenario

Federated multi-view clustering is designed to collaboratively mine heterogeneous multi-source information across clients. However, existing methods typically assume uniform view distributions across clients, thereby overlooking the dual uncertainties of view uncertainty (semantic inconsistency arising from arbitrary pairings of views) and aggregation uncertainty (divergent update directions and imbalanced contributions among clients). To address these, we propose a novel Enhanced Federated Deep Multi-View Clustering framework: hierarchical contrastive alignment within clients resolves view uncertainty by eliminating semantic conflicts; a view-adaptive drift module mitigates aggregation uncertainty through global-local prototype contrast that dynamically corrects parameter deviations; and a contribution-aware aggregation mechanism coordinates client updates. Experimental results demonstrate that EFDMVC achieves superior robustness against heterogeneous uncertain views across multiple benchmark datasets, consistently outperforming all state-of-the-art baselines in comprehensive evaluations.