ParBalans: Parallel Multi-Armed Bandits-based Adaptive Large Neighborhood Search

2025-08-08

AI Summary

To address the low efficiency and poor scalability of solving large-scale mixed-integer programming (MIP) problems, this paper proposes ParBalans—an adaptive large neighborhood search (ALNS) framework integrating solver-level and algorithm-level parallelism. Building upon Balans, ParBalans innovatively employs multi-armed bandits (MAB) to guide online, adaptive exploration of parallel parameter configurations, enabling cooperative optimization of search strategies. Its modular architecture and fine-grained task scheduling ensure robustness while significantly improving parallel speedup. Experimental evaluation on challenging MIP benchmark instances demonstrates that ParBalans achieves solution times competitive with state-of-the-art commercial solvers such as Gurobi, attaining an average speedup of 2.3× over sequential Balans. These results validate the effectiveness and scalability of MAB-driven parallel metaheuristics for MIP solving.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Multiagent Systems: Distributed Problem Solving
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Solving Mixed-Integer Programming (MIP) problems often requires substantial computational resources due to their combinatorial nature. Parallelization has emerged as a critical strategy to accelerate solution times and enhance scalability to tackle large, complex instances. This paper investigates the parallelization capabilities of Balans, a recently proposed multi-armed bandits-based adaptive large neighborhood search for MIPs. While Balans's modular architecture inherently supports parallel exploration of diverse parameter configurations, this potential has not been thoroughly examined. To address this gap, we introduce ParBalans, an extension that leverages both solver-level and algorithmic-level parallelism to improve performance on challenging MIP instances. Our experimental results demonstrate that ParBalans exhibits competitive performance compared to the state-of-the-art commercial solver Gurobi, particularly on hard optimization benchmarks.

Problem

• Accelerates Mixed-Integer Programming via parallelization
• Enhances scalability for large complex MIP instances
• Improves performance using solver and algorithmic parallelism

Innovation

• Parallel multi-armed bandits-based adaptive search
• Solver and algorithmic-level parallelism integration
• Modular architecture for diverse parameter exploration