Partial multi-label feature selection via low-rank and sparse factorization with manifold learning

Zhenzhen Sun, Zexiang Chen, Jinghua Liu, Yewang Chen, Yuanlong Yu

Abstract: Highlights•A novel feature selection approach is proposed for partial multi-label learning.•A low-rank and sparse factorization model is designed to disambiguate noisy labels.•A feature graph is designed to preserve the local label correlations.•An efficient optimization method with provable convergence is designed.