Learning to Solve the Quadratic Assignment Problem with Warm-Started MCMC Finetuning

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Pan, Yicheng, Zhou, Ruisong, Zou, Haijun, Li, Tianyou, Wen, Zaiwen
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866910156897386496
author Pan, Yicheng
Zhou, Ruisong
Zou, Haijun
Li, Tianyou
Wen, Zaiwen
author_facet Pan, Yicheng
Zhou, Ruisong
Zou, Haijun
Li, Tianyou
Wen, Zaiwen
contents The quadratic assignment problem (QAP) is a fundamental NP-hard task that poses significant challenges for both traditional heuristics and modern learning-based solvers. Existing QAP solvers still struggle to achieve consistently competitive performance across structurally diverse real-world instances. To bridge this performance gap, we propose PLMA, an innovative permutation learning framework. PLMA features an efficient warm-started MCMC finetuning procedure to enhance deployment-time performance, leveraging short Markov chains to anchor the adaptation to the promising regions previously explored. For rapid exploration via MCMC over the permutation space, we design an additive energy-based model (EBM) that enables an $O(1)$-time 2-swap Metropolis-Hastings sampling step. Moreover, the neural network used to parameterize the EBM incorporates a scalable and flexible cross-graph attention mechanism to model interactions between facilities and locations in the QAP. Extensive experiments demonstrate that PLMA consistently outperforms state-of-the-art baselines across various benchmarks. In particular, PLMA achieves a near-zero average optimality gap on QAPLIB, exhibits remarkably superior robustness on the notoriously difficult Taixxeyy instances, and also serves as an effective QAP solver in bandwidth minimization.
format Preprint
id arxiv_https___arxiv_org_abs_2604_20109
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Learning to Solve the Quadratic Assignment Problem with Warm-Started MCMC Finetuning
Pan, Yicheng
Zhou, Ruisong
Zou, Haijun
Li, Tianyou
Wen, Zaiwen
Machine Learning
Artificial Intelligence
Optimization and Control
90C27, 68T20
The quadratic assignment problem (QAP) is a fundamental NP-hard task that poses significant challenges for both traditional heuristics and modern learning-based solvers. Existing QAP solvers still struggle to achieve consistently competitive performance across structurally diverse real-world instances. To bridge this performance gap, we propose PLMA, an innovative permutation learning framework. PLMA features an efficient warm-started MCMC finetuning procedure to enhance deployment-time performance, leveraging short Markov chains to anchor the adaptation to the promising regions previously explored. For rapid exploration via MCMC over the permutation space, we design an additive energy-based model (EBM) that enables an $O(1)$-time 2-swap Metropolis-Hastings sampling step. Moreover, the neural network used to parameterize the EBM incorporates a scalable and flexible cross-graph attention mechanism to model interactions between facilities and locations in the QAP. Extensive experiments demonstrate that PLMA consistently outperforms state-of-the-art baselines across various benchmarks. In particular, PLMA achieves a near-zero average optimality gap on QAPLIB, exhibits remarkably superior robustness on the notoriously difficult Taixxeyy instances, and also serves as an effective QAP solver in bandwidth minimization.
title Learning to Solve the Quadratic Assignment Problem with Warm-Started MCMC Finetuning
topic Machine Learning
Artificial Intelligence
Optimization and Control
90C27, 68T20
url https://arxiv.org/abs/2604.20109