Comparison between the Iterative Local Search and Exhaustive Search methods applied to QAOA in Max-Cut and Ising Spin Model problems

TL;DR

This paper compares Exhaustive Search and Iterative Local Search methods for optimizing QAOA applied to Max-Cut and Ising Spin Model problems, evaluating their effectiveness in classical and quantum simulations.

Contribution

It provides a comparative analysis of ES and ILS techniques for QAOA in different problem configurations on classical and quantum hardware.

Findings

ILS outperforms ES in certain configurations

Quantum simulations show promising results for QAOA optimization

Classical approaches help understand QAOA's potential in real quantum hardware

Abstract

A comparison is made between Exhaustive Search (ES) and Iterative Local Search (ILS). Such comparison was made using the Quantum Approximation Optimization Algorithm (QAOA). QAOA has been extensively researched due to its this potential to be implemented in actual quantum hardware, and its promising future in optimization problems and quantum machine learning. ES and ILS approaches were simulated to determine the pros and cons of these techniques for QAOA in local (classic computer) and real simulations (IBM quantum computer). These classic approaches were used in QAOA to approximate the optimal expected value in Max-Cut and Ising Spin Model (ISM) problems, both of these flavors have three simulated configurations called: linear, cyclic and complete (or full).