Abstract: Machine learning has demonstrated remarkable effectiveness in solving scheduling problems through end-to-end optimization. However, dynamic events introduce uncertainty and pose significant ...