Teng Pang

Matrix-wise Class Imbalance Matters: On the Generalization of Micro-AUC in Multi-label Learning