Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
80 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

MVMoE: Multi-Task Vehicle Routing Solver with Mixture-of-Experts (2405.01029v2)

Published 2 May 2024 in cs.AI and cs.LG

Abstract: Learning to solve vehicle routing problems (VRPs) has garnered much attention. However, most neural solvers are only structured and trained independently on a specific problem, making them less generic and practical. In this paper, we aim to develop a unified neural solver that can cope with a range of VRP variants simultaneously. Specifically, we propose a multi-task vehicle routing solver with mixture-of-experts (MVMoE), which greatly enhances the model capacity without a proportional increase in computation. We further develop a hierarchical gating mechanism for the MVMoE, delivering a good trade-off between empirical performance and computational complexity. Experimentally, our method significantly promotes zero-shot generalization performance on 10 unseen VRP variants, and showcases decent results on the few-shot setting and real-world benchmark instances. We further conduct extensive studies on the effect of MoE configurations in solving VRPs, and observe the superiority of hierarchical gating when facing out-of-distribution data. The source code is available at: https://github.com/RoyalSkye/Routing-MVMoE.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (88)
  1. Neural combinatorial optimization with reinforcement learning. In ICLR Workshop Track, 2017.
  2. Machine learning for combinatorial optimization: a methodological tour d’horizon. European Journal of Operational Research, 290(2):405–421, 2021.
  3. RL4CO: a unified reinforcement learning for combinatorial optimization library. In NeurIPS 2023 Workshop: New Frontiers in Graph Learning, 2023.
  4. Learning generalizable models for vehicle routing problems via knowledge distillation. In NeurIPS, 2022.
  5. Machine learning to solve vehicle routing problems: A survey. IEEE Transactions on Intelligent Transportation Systems, 2024.
  6. Towards a generic representation of cominatorial problems for learning-based approaches. arXiv preprint arXiv:2403.06026, 2024.
  7. Vehicle routing problems for city logistics. EURO Journal on Transportation and Logistics, 6(1):51–79, 2017.
  8. Combinatorial optimization with policy adaptation using latent space search. In NeurIPS, 2023.
  9. Neural multi-objective combinatorial optimization with diversity enhancement. In NeurIPS, 2023.
  10. Learning to perform local rewriting for combinatorial optimization. In NeurIPS, volume 32, 2019.
  11. Towards understanding the mixture-of-experts layer in deep learning. In NeurIPS, volume 35, pp.  23049–23062, 2022.
  12. Croes, G. A. A method for solving traveling-salesman problems. Operations research, 6(6):791–812, 1958.
  13. Learning 2-opt heuristics for the traveling salesman problem via deep reinforcement learning. In Asian Conference on Machine Learning, pp.  465–480. PMLR, 2020.
  14. BQ-NCO: Bisimulation quotienting for generalizable neural combinatorial optimization. In NeurIPS, 2023.
  15. Learning factored representations in a deep mixture of experts. arXiv preprint arXiv:1312.4314, 2013.
  16. A review of sparse expert models in deep learning. arXiv preprint arXiv:2209.01667, 2022a.
  17. Switch transformers: Scaling to trillion parameter models with simple and efficient sparsity. The Journal of Machine Learning Research, 23(1):5232–5270, 2022b.
  18. Gpt-3: Its nature, scope, limits, and consequences. Minds and Machines, 30:681–694, 2020.
  19. Generalize a small pre-trained model to arbitrarily large tsp instances. In AAAI, volume 35, pp.  7474–7482, 2021.
  20. Or-tools routing library, 2023. URL https://developers.google.com/optimization/routing.
  21. Towards generalizable neural solvers for vehicle routing problems via ensemble with transferrable local policy. arXiv preprint arXiv:2308.14104, 2023.
  22. Generalization of neural combinatorial solvers through the lens of adversarial robustness. In ICLR, 2022.
  23. Winner takes it all: Training performant RL populations for combinatorial optimization. In NeurIPS, 2023.
  24. Helsgaun, K. An extension of the lin-kernighan-helsgaun tsp solver for constrained traveling salesman and vehicle routing problems. Roskilde: Roskilde University, pp.  24–50, 2017.
  25. Neural large neighborhood search for the capacitated vehicle routing problem. In European Conference on Artificial Intelligence, pp.  443–450, 2020.
  26. PolyNet: Learning diverse solution strategies for neural combinatorial optimization. arXiv preprint arXiv:2402.14048, 2024.
  27. Generalize learned heuristics to solve large-scale vehicle routing problems in real-time. In ICLR, 2023.
  28. LoRA: Low-rank adaptation of large language models. In ICLR, 2022.
  29. Graph neural network guided local search for the traveling salesperson problem. In ICLR, 2022.
  30. Interpretable mixture of experts. Transactions on Machine Learning Research, 2023. ISSN 2835-8856.
  31. Adaptive mixtures of local experts. Neural computation, 3(1):79–87, 1991.
  32. Ensemble-based deep reinforcement learning for vehicle routing problems under distribution shift. In NeurIPS, 2023.
  33. Hierarchical mixtures of experts and the em algorithm. Neural computation, 6(2):181–214, 1994.
  34. An efficient graph convolutional network technique for the travelling salesman problem. arXiv preprint arXiv:1906.01227, 2019.
  35. Learning tsp requires rethinking generalization. In International Conference on Principles and Practice of Constraint Programming, 2021.
  36. Scaling laws for neural language models. arXiv preprint arXiv:2001.08361, 2020.
  37. Sym-NCO: Leveraging symmetricity for neural combinatorial optimization. In NeurIPS, 2022.
  38. Ant colony sampling with gflownets for combinatorial optimization. arXiv preprint arXiv:2403.07041, 2024.
  39. Attention, learn to solve routing problems! In ICLR, 2018.
  40. Deep policy dynamic programming for vehicle routing problems. In International conference on integration of constraint programming, artificial intelligence, and operations research, pp.  190–213, 2022.
  41. Scaling laws for fine-grained mixture of experts. arXiv preprint arXiv:2402.07871, 2024.
  42. POMO: Policy optimization with multiple optima for reinforcement learning. In NeurIPS, volume 33, pp.  21188–21198, 2020.
  43. Matrix encoding networks for neural combinatorial optimization. In NeurIPS, volume 34, pp.  5138–5149, 2021.
  44. Gshard: Scaling giant models with conditional computation and automatic sharding. arXiv preprint arXiv:2006.16668, 2020.
  45. Base layers: Simplifying training of large, sparse models. In ICML, pp.  6265–6274. PMLR, 2021.
  46. Deep reinforcement learning for solving the heterogeneous capacitated vehicle routing problem. IEEE Transactions on Cybernetics, 52(12):13572–13585, 2021a.
  47. Learning to delegate for large-scale vehicle routing. In NeurIPS, volume 34, pp.  26198–26211, 2021b.
  48. Prefix-tuning: Optimizing continuous prompts for generation. In ACL, pp.  4582–4597, 2021.
  49. Cross-problem learning for solving vehicle routing problems. In IJCAI, 2024.
  50. Multi-task learning for routing problem with cross-problem zero-shot generalization. arXiv preprint arXiv:2402.16891, 2024.
  51. A learning-based iterative method for solving vehicle routing problems. In ICLR, 2020.
  52. Neural combinatorial optimization with heavy decoder: Toward large scale generalization. In NeurIPS, 2023.
  53. Learning to search feasible and infeasible regions of routing problems with flexible neural k-opt. In NeurIPS, 2023.
  54. On the generalization of neural combinatorial optimization heuristics. In ECML PKDD, 2022.
  55. Unsupervised learning for solving the travelling salesman problem. In NeurIPS, 2023.
  56. Reinforcement learning for solving the vehicle routing problem. In NeurIPS, volume 31, 2018.
  57. Demystifying softmax gating function in gaussian mixture of experts. In NeurIPS, 2023.
  58. Statistical perspective of top-k sparse softmax gating mixture of experts. In ICLR, 2024.
  59. From sparse to soft mixtures of experts. In ICLR, 2024.
  60. DIMES: A differentiable meta solver for combinatorial optimization problems. In NeurIPS, 2022.
  61. Scaling vision with sparse mixture of experts. In NeurIPS, volume 34, pp.  8583–8595, 2021.
  62. Hash layers for large sparse models. In NeurIPS, volume 34, pp.  17555–17566, 2021.
  63. Independent prototype propagation for zero-shot compositionality. NeurIPS, 34:10641–10653, 2021.
  64. Shaw, P. Using constraint programming and local search methods to solve vehicle routing problems. In International conference on principles and practice of constraint programming, pp.  417–431. Springer, 1998.
  65. Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. In ICLR, 2017.
  66. Solomon, M. M. Algorithms for the vehicle routing and scheduling problems with time window constraints. Operations research, 35(2):254–265, 1987.
  67. Meta-SAGE: Scale meta-learning scheduled adaptation with guided exploration for mitigating scale shift on combinatorial optimization. In ICML, 2023.
  68. DIFUSCO: Graph-based diffusion solvers for combinatorial optimization. In NeurIPS, 2023.
  69. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288, 2023.
  70. New benchmark instances for the capacitated vehicle routing problem. European Journal of Operational Research, 257(3):845–858, 2017.
  71. Attention is all you need. In NeurIPS, volume 30, 2017.
  72. Vidal, T. Hybrid genetic search for the cvrp: Open-source implementation and swap* neighborhood. Computers & Operations Research, 140:105643, 2022.
  73. Pointer networks. In NeurIPS, volume 28, 2015.
  74. Efficient training of multi-task neural solver with multi-armed bandits. arXiv preprint arXiv:2305.06361, 2023.
  75. ASP: Learn a universal neural solver! IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024.
  76. Williams, R. J. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning, 8(3):229–256, 1992.
  77. Learning improvement heuristics for solving routing problems. IEEE transactions on neural networks and learning systems, 33(9):5057–5069, 2021.
  78. Neural airport ground handling. IEEE Transactions on Intelligent Transportation Systems, 2023.
  79. NeuroLKH: Combining deep learning model with lin-kernighan-helsgaun heuristic for solving the traveling salesman problem. In NeurIPS, volume 34, pp.  7472–7483, 2021.
  80. OpenMoE: An early effort on open mixture-of-experts language models. arXiv preprint arXiv:2402.01739, 2024.
  81. DeepACO: Neural-enhanced ant systems for combinatorial optimization. In NeurIPS, 2023.
  82. Twenty years of mixture of experts. IEEE transactions on neural networks and learning systems, 23(8):1177–1193, 2012.
  83. A review on learning to solve combinatorial optimisation problems in manufacturing. IET Collaborative Intelligent Manufacturing, 5(1):e12072, 2023.
  84. Learning to solve travelling salesman problem with hardness-adaptive curriculum. In AAAI, 2022.
  85. Learning large neighborhood search for vehicle routing in airport ground handling. IEEE Transactions on Knowledge and Data Engineering, 2023a.
  86. Towards omni-generalizable neural methods for vehicle routing problems. In ICML, pp.  42769–42789. PMLR, 2023b.
  87. Mixture-of-experts with expert choice routing. In NeurIPS, volume 35, pp.  7103–7114, 2022.
  88. Taming sparsely activated transformer with stochastic experts. In ICLR, 2022.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (7)
  1. Jianan Zhou (13 papers)
  2. Zhiguang Cao (48 papers)
  3. Yaoxin Wu (26 papers)
  4. Wen Song (24 papers)
  5. Yining Ma (31 papers)
  6. Jie Zhang (846 papers)
  7. Chi Xu (54 papers)
Citations (11)
View on Semantic Scholar
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets