【學術活動】Deep Reinforcement Learning for Combinatorial Optimization

Abstract: This lecture will present recent advances in applying artificial intelligence and machine learning techniques to combinatorial optimization problems, with a particular focus on vehicle routing problems (VRPs). The talk will introduce hybrid optimization frameworks that integrate deep reinforcement learning, metaheuristics, and mathematical optimization to address complex logistics and transportation challenges. Real-world applications and case studies from both industry and academia will illustrate how data-driven approaches can improve routing efficiency, reduce operational costs, and support sustainable logistics operations. The lecture will also discuss the challenges of handling dynamic and uncertain environments, including fluctuating customer demands, traffic congestion, and delivery time constraints, as well as future research directions for developing scalable, adaptive, and environmentally sustainable optimization systems.

Date & Time: 4:30 PM–6:00 PM, Monday, May 25, 2026

Location: MA207, NTUST