Efficient Co-clustering via Anchor-refined Label Spreading

Anchor graph clustering has gained significant attention due to its effectiveness and efficiency. As the most representative points in original data, anchors are applied to connect the sample space to label space. However, when noise is present in original data, the anchor-refined label spreading mechanism may fail. To address this, we propose an Efficient Co-clustering via Anchor-refined Label Spreading (ECALS), which simultaneously clusters original data and anchors. We introduce the size constraint that ensures each cluster contains a minimum number of samples. Our method includes two variants, which are continuous and discrete model, catering to both fuzzy and discrete label matrices. Both models are applicable to out-of-sample problems and demonstrate superior performance on synthetic and real-world datasets.