机器人学研究室知识库

Multi-view outlier detection has been well investigated in recent years. However, 1) most existing methods cannot efficiently handle outlier detection problem for large-scale multi-view data, since exploring pairwise constraints among different views causes highly-computational cost; 2) the data collected from heterogeneous feature spaces further increases the difficulty of multi-view outlier detection. To address these issues, we present a fast multi-view outlier detection model via learning a low-rank latent subspace representation with deep encoder architecture, which can not only identify the outliers for large-scale data even with numerous data views, but also exploit a common latent subspace shared by all views. First, we learn a view-specific dictionaries from a small dataset sampled from original dataset. Benefiting from view-specific dictionaries, the sampled data is projected as a shared and discriminative latent representations, which correspond to view-consistent and view-specific components across multiple views, respectively. Then, the obtained discriminative representations are applied to train the view-specific deep encoders, which can efficiently compute the abnormal score for the remaining instances. Our model can cost-effectively identify outliers in large-scale datasets from numerous data views with less computational complexity. Experiments conducted on eight datasets and a synthesis dataset show that our model outperforms existing ones effectively.