Multi-Task Learning for Routing Problem with Cross-Problem Zero-Shot Generalization

TL;DR

This paper introduces a unified neural model for vehicle routing problems that generalizes across different problem types without retraining, significantly improving performance over existing methods in diverse scenarios.

Contribution

The work presents the first approach to cross-problem generalization in VRPs using attribute composition, enabling a single model to solve multiple variants with unseen attribute combinations.

Findings

Achieved around 5% average gap on multiple VRP variants.

Reduced performance gap from over 20% to about 5%.

Demonstrated effectiveness on real-world logistics scenarios.

Abstract

Vehicle routing problems (VRPs), which can be found in numerous real-world applications, have been an important research topic for several decades. Recently, the neural combinatorial optimization (NCO) approach that leverages a learning-based model to solve VRPs without manual algorithm design has gained substantial attention. However, current NCO methods typically require building one model for each routing problem, which significantly hinders their practical application for real-world industry problems with diverse attributes. In this work, we make the first attempt to tackle the crucial challenge of cross-problem generalization. In particular, we formulate VRPs as different combinations of a set of shared underlying attributes and solve them simultaneously via a single model through attribute composition. In this way, our proposed model can successfully solve VRPs with unseen attribute combinations in a zero-shot generalization manner. Extensive experiments are conducted on eleven VRP variants, benchmark datasets, and industry logistic scenarios. The results show that the unified model demonstrates superior performance in the eleven VRPs, reducing the average gap to around 5% from over 20% in the existing approach and achieving a significant performance boost on benchmark datasets as well as a real-world logistics application. The source code is included in https://github.com/FeiLiu36/MTNCO.