让神经路由模型提前看多步，提升长期规划能力
Learning with Foresight: Enhancing Neural Routing Policy via Multi-Node Lookahead Prediction
- 训练时预测多个未来节点，突破只看下一步的局限
- 在不同规模和真实场景下，性能显著优于现有方法
- 可无缝适配各类模型，推理时无额外开销
神经策略在求解车辆路径问题上展现潜力，因其减少了对人工设计启发式规则的依赖。然而，当前训练范式存在根本缺陷：主要聚焦于构建解时的下一步节点预测，导致决策短视，削弱了长程规划能力。为此，我们提出多节点前瞻预测（MnLP），一种新颖的训练策略，将监督学习扩展为同时预测多个未来节点。引入因果性与可丢弃的MnLP模块，仅在训练阶段运行，帮助模型预判多步决策，同时保持推理效率。通过在损失函数中加入多深度辅助监督，MnLP使神经策略具备长程上下文理解能力。实验表明，MnLP优于现有训练方法，在不同问题规模、分布及真实基准上均提升神经策略的泛化能力。此外，该方法可无缝集成至多种神经架构，且不增加推理开销。
原文摘要 · Abstract (English)
Neural policies have shown promise in solving vehicle routing problems due to their reduced reliance on handcrafted heuristics. However, current training paradigms suffer from a fundamental limitation: they primarily focus on next-node prediction for solution construction, resulting in myopic decision-making that undermines long-horizon planning capacity. To this end, we introduce Multi-node Lookahead Prediction (MnLP), a novel training strategy that extends the supervised learning paradigm to predict multiple future nodes simultaneously. We incorporate causal and discardable MnLP modules that operate exclusively during training, facilitating models to anticipate multi-step decisions while preserving inference-time efficiency. By incorporating multi-depth auxiliary supervision into the loss function, MnLP equips neural policies with the ability of long-range contextual understanding. Experimentally, MnLP outperforms existing training methods, improving the generalization capability of neural policies across various problem sizes, distributions, and real-world benchmarks. Moreover, MnLP can be seamlessly integrated into diverse neural architectures without introducing additional inference overhead.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。