A Study on the Leapfrogging Strategy for the Quantum Approximate Optimization Algorithm on the Max-cut of n-regular Graph Instances

摘要

The quantum approximate optimization algorithm(QAOA) has numerous promising applications onsolving the combinatorial optimization problems onthe near-term Noisy Intermediate Scalable Quantum(NISQ) devices. QAOA has a quantum-classical hybridstructure, with the quantum part consisting theparameterized alternating operator ansatz, and theclassical part consist of an optimization algorithm optimizingthe parameters to maximize the expectationvalue. This value depends highly on the parameters.This implies that a set of good parameters leads toan accurate solution of the given problem. However,at large circuit depth, it is dicult to achieve globaloptimization due to the multiple occurrence of localmaxima. Therefore, we study the so-called leapfroggingstrategy on solving the Max-cut problem for 3-regular graphs, which reuses the optimized parametersin larger graphs.