Hierarchical Quantum Architecture Search for Variational Quantum Algorithms

Abstract

Designing efficient variational quantum algorithms (VQAs) is crucial for transforming the theoretical advantages of quantum algorithms into practical applications. In this context, quantum architecture search (QAS) has been introduced to automate the search and design of VQAs. However, current mainstream QAS algorithms typically perform both global and local searches simultaneously, which can result in high search space complexity and optimization challenges. In this paper, we propose a hierarchical quantum architecture search framework based on a two-stage search structure. In the first stage, global exploration of the overall quantum circuit structure is performed, while in the second stage, local optimization of quantum gate selection is carried out. We provide a numerical analysis of the theoretical advantages of the proposed framework in reducing the search space. To evaluate practical performance, we conduct experiments on quantum chemistry tasks with different algorithm combinations integrated into the framework. The results demonstrate the effectiveness of the hierarchical search structure in automating quantum circuit design.

Keywords

• Automated quantum circuit design
• hierarchical quantum architecture search