Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

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

About

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.

Yicheng Pan, Ruisong Zhou, Haijun Zou, Tianyou Li, Zaiwen Wen• 2026

Related benchmarks

TaskDatasetResultRank
Bandwidth MinimizationUF Sparse Matrix Collection (test)
Bandwidth4
136
Quadratic Assignment ProblemQAPLIB
Gap0.00e+0
80
Quadratic Assignment ProblemTaixxeyy
Mean Gap (%)0.00e+0
24
Quadratic Assignment ProblemGeometrically Structured synthetic datasets n=100
Solution Cost1.59e+3
13
Quadratic Assignment ProblemGeometrically Structured synthetic datasets (n=50)
Cost375.5
13
Quadratic Assignment ProblemGeometrically Structured synthetic datasets n=20
Solution Cost54.37
13
Quadratic Assignment ProblemUniformly Random synthetic datasets n=50
Cost521.8
12
Quadratic Assignment ProblemUniformly Random synthetic datasets n=100
Cost2.19e+3
12
Quadratic Assignment ProblemUniformly Random synthetic datasets n=20
Cost76.56
12
Quadratic Assignment ProblemLarge-scale synthetic QAP dataset QAP200
Cost6.64e+3
5
Showing 10 of 11 rows

Other info

Follow for update