Pointer Networks with Q-Learning for OP Combinatorial Optimization

arXiv:2311.02629v2 Announce Type: replace
Abstract: The Orienteering Problem (OP) presents a unique challenge in Combinatorial Optimization (CO), emphasized by its widespread use in logistics, delivery, and transportation planning. Given the NP-hard nature of OP, obtaining optimal solutions is inherently complex. While Pointer Networks (Ptr-Nets) have exhibited prowess in various combinatorial tasks, their performance in the context of OP, and duties requiring focus on future return or exploration, leaves room for improvement. Recognizing the potency combining Reinforcement Learning (RL) methods with …

abstract arxiv challenge cs.lg delivery logistics math.oc nature networks np-hard optimization planning q-learning solutions tasks transportation type