Abstract: Highlights•Proposed approach deals with class imbalance issue in multi-label learning.•Unified framework for feature selection and multi-label classification.•Learning Label Specific Features.•Incorporating structural property of data and label correlation.
Loading