Robust Multi-view Clustering against Imperfect Information
Researchers have developed a new multi-view clustering framework called PLCI to address challenges posed by imperfect information in real-world datasets. This framework unifies the handling of incomplete views and noisy correspondences by treating cross-view counterparts as latent variables. PLCI integrates instance-level reliability and semantic transport to infer the posterior distribution of these latent counterparts, demonstrating effectiveness across six datasets against ten state-of-the-art methods. AI