MG-Net: Learn to Customize QAOA with Circuit Depth Awareness

TL;DR

This paper introduces MG-Net, a deep learning framework that customizes QAOA mixer Hamiltonians based on circuit depth, improving optimization performance on combinatorial problems with practical quantum device constraints.

Contribution

The paper presents MG-Net, a novel deep learning approach that dynamically generates problem-specific mixer Hamiltonians for QAOA, addressing circuit depth limitations.

Findings

MG-Net outperforms traditional methods in approximation ratio.

Systematic simulations validate MG-Net's effectiveness on large-scale problems.

Theoretical analysis links mixer design, problem type, and circuit depth.

Abstract

Quantum Approximate Optimization Algorithm (QAOA) and its variants exhibit immense potential in tackling combinatorial optimization challenges. However, their practical realization confronts a dilemma: the requisite circuit depth for satisfactory performance is problem-specific and often exceeds the maximum capability of current quantum devices. To address this dilemma, here we first analyze the convergence behavior of QAOA, uncovering the origins of this dilemma and elucidating the intricate relationship between the employed mixer Hamiltonian, the specific problem at hand, and the permissible maximum circuit depth. Harnessing this understanding, we introduce the Mixer Generator Network (MG-Net), a unified deep learning framework adept at dynamically formulating optimal mixer Hamiltonians tailored to distinct tasks and circuit depths. Systematic simulations, encompassing Ising models and weighted Max-Cut instances with up to 64 qubits, substantiate our theoretical findings, highlighting MG-Net's superior performance in terms of both approximation ratio and efficiency.