Multi-view classifier based on Probabilistic Collaborative Representation and Latent Representation

Multi-view is quite more effective at improving the training of model than merely using single view. However, most existing multi-view learning algorithms only either pay attention to consistency or complementary principle among views, not making full use of multi-view data. Due to its high complexity, algorithm considering both complementarity and consistency has limited ability to process large-scale data. On the basis of Probabilistic Collaborative Representa-tion based Classifier (ProCRC), we propose Probabilistic Collaborative Representation based Classifier for Multi-View (ProCRC-MV), which jointly maximizes the likelihood that a test example belongs to the co-subspace of each class. Learning subspace in the process of collaborative representation, considering consistency and complementarity concur-rently, ProCRC-MV can achieve promising classification performance. Meanwhile, it has low computational complexity, fast running speed, and can still maintain good performance when dealing with large-scale data. ProCRC-MV has the ability for subspace learning based on self-representation, so we combain latent representation learning for better search-ing subspace with ProCRC-MV to construct a novel classifier called LProCRC-MV, the ability of LProCRC-MV to process complex data is further enhanced comparing with ProCRC-MV.