Reusability Report: Comparing gradient descent and monte carlo tree search optimization of quantum annealing schedules

TL;DR

This paper evaluates and compares the reusability of gradient descent and Monte Carlo tree search methods for optimizing quantum annealing schedules, with additional benchmarks on Max-Cut problems to assess their effectiveness.

Contribution

It provides a reusability report of a recent quantum annealing schedule optimization method and extends the analysis with new benchmarks on Max-Cut problems.

Findings

Monte Carlo tree search shows promising results in schedule optimization.

Gradient descent offers competitive performance in certain problem instances.

The benchmarks highlight the strengths and limitations of each method.

Abstract

We provide a reusability report of the method presented by Chen et al. in "Optimizing quantum annealing schedules with Monte Carlo tree search enhanced with neural networks" and add further benchmarks on Max-Cut problems.