A Unified Framework for Graph-Based Multi-View Partial Multi-Label Learning

Multi-view partial multi-label learning (MVPML) is a fundenmental problem where each sample linked to multiple kinds of features and candidate labels, including ground-truth noise labels. The key MVPML how manipulate the recover labels from label set. To this end, study designs novel Graph-based Partial Multi-label model named as GMPM, which combines multi-view information detection, valuable selection predictor into unified optimization model. be specific, GMPM first exploits consensus across views by view-specific similarity graph fuses graphs target one. Then, we divide observed set two parts: part part, latter associated with sparse constraint make sure former clean. Furthermore, embed learned process disambiguation restore more reliable matrix. Finally, resulting predictive help Extensive experiments on six common used datasets demonstrate that proposed achieves comparable performance over state-of-the-arts.