Learning to Solve Routing Problems via Distributionally Robust Optimization
This addresses the issue of limited generalization in routing algorithms for researchers and practitioners, though it is incremental as it builds on existing deep models.
The paper tackled the problem of poor cross-distribution generalization in deep models for routing problems by using group distributionally robust optimization, resulting in significant improvements in performance on benchmark datasets like TSPLib and CVRPLib.
Recent deep models for solving routing problems always assume a single distribution of nodes for training, which severely impairs their cross-distribution generalization ability. In this paper, we exploit group distributionally robust optimization (group DRO) to tackle this issue, where we jointly optimize the weights for different groups of distributions and the parameters for the deep model in an interleaved manner during training. We also design a module based on convolutional neural network, which allows the deep model to learn more informative latent pattern among the nodes. We evaluate the proposed approach on two types of well-known deep models including GCN and POMO. The experimental results on the randomly synthesized instances and the ones from two benchmark dataset (i.e., TSPLib and CVRPLib) demonstrate that our approach could significantly improve the cross-distribution generalization performance over the original models.