Enhancing the Cross-Size Generalization for Solving Vehicle Routing Problems via Continual Learning

Jingwen Li, Zhiguang Cao, Yaoxin Wu, Tang Liu

Primary Area: optimization

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: Vehicle routing problem, generalization, continual learning

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

Abstract: Deep models for vehicle routing problems are typically trained and evaluated using instances of a single size, which severely limits their ability to generalize across different problem sizes and thus hampers their practical applicability. To address the issue, we propose a continual learning based framework that sequentially trains a deep model with instances of ascending problem sizes. Specifically, on the one hand, we design an inter-task regularization scheme to retain the knowledge acquired from smaller problem sizes in the model training on a larger size. On the other hand, we introduce an intra-task regularization scheme to consolidate the model by imitating the latest desirable behaviors during training on each size. Additionally, we exploit the experience replay to revisit instances of formerly trained sizes for mitigating the catastrophic forgetting. Extensive experimental results show that the proposed approach achieves predominantly superior performance across various problem sizes (either seen or unseen in the training), as compared to state-of-the-art deep models including the ones specialized for the generalizability enhancement. Meanwhile, the ablation studies on the key designs manifest their synergistic effect in the proposed framework.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 5661