QAOA Parameter Transferability for Maximum Independent Set using Graph Attention Networks

2025-04-29

AI Summary

Solving the Maximum Independent Set (MIS) problem on large-scale graphs using Noisy Intermediate-Scale Quantum (NISQ) devices remains challenging due to prohibitive parameter optimization costs in the Quantum Approximate Optimization Algorithm (QAOA). Method: We propose a cross-scale QAOA parameter transfer framework leveraging Graph Attention Networks (GATs), where a GAT trained on small graphs efficiently predicts high-quality initial parameters for large graphs (up to thousands of vertices); this is integrated into HyDRA-MIS—a distributed, resource-aware framework combining graph decomposition, variational quantum optimization, and parallel execution. Contribution/Results: This work pioneers the use of GATs for QAOA parameter generalization across graph scales. HyDRA-MIS significantly reduces parameter optimization iterations and hardware resource consumption while preserving solution quality. Experiments demonstrate performance on par with the state-of-the-art classical solver KaMIS on large graphs, establishing a scalable quantum-classical co-design paradigm for combinatorial optimization in the NISQ era.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Machine Learning: Quantum Machine Learning
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

The quantum approximate optimization algorithm (QAOA) is one of the promising variational approaches of quantum computing to solve combinatorial optimization problems. In QAOA, variational parameters need to be optimized by solving a series of nonlinear, nonconvex optimization programs. In this work, we propose a QAOA parameter transfer scheme using Graph Attention Networks (GAT) to solve Maximum Independent Set (MIS) problems. We prepare optimized parameters for graphs of 12 and 14 vertices and use GATs to transfer their parameters to larger graphs. Additionally, we design a hybrid distributed resource-aware algorithm for MIS (HyDRA-MIS), which decomposes large problems into smaller ones that can fit onto noisy intermediate-scale quantum (NISQ) computers. We integrate our GAT-based parameter transfer approach to HyDRA-MIS and demonstrate competitive results compared to KaMIS, a state-of-the-art classical MIS solver, on graphs with several thousands vertices.

Problem

• Transfer QAOA parameters for MIS using Graph Attention Networks
• Optimize variational parameters for quantum combinatorial optimization
• Decompose large MIS problems for NISQ quantum computers

Innovation

• QAOA parameter transfer using Graph Attention Networks
• Hybrid distributed algorithm for large MIS problems
• Parameter transfer from small to large graphs