Creating Automated Quantum-Assisted Solutions for Optimization Problems

TL;DR

The paper introduces the QuaST decision tree framework, which streamlines the process of applying quantum-assisted optimization solutions by providing a structured, modular approach to explore, automate, and evaluate different solution paths for real-world problems.

Contribution

It presents a novel, flexible decision tree framework that supports end users and researchers in systematically developing quantum-assisted optimization solutions.

Findings

Framework enables exploration of multiple solution paths

Supports transfer of research to practical applications

Facilitates gathering experience with real-world use cases

Abstract

When trying to use quantum-enhanced methods for optimization problems, the sheer number of options inhibits its adoption by industrial end users. Expert knowledge is required for the formulation and encoding of the use case, the selection and adaptation of the algorithm, and the identification of a suitable quantum computing backend. Navigating the decision tree spanned by these options is a difficult task and supporting integrated tools are still missing. We propose the QuaST decision tree, a framework that allows to explore, automate and systematically evaluate solution paths. It helps end users to transfer research to their application area, and researchers to gather experience with real-world use cases. Our setup is modular, highly structured and flexible enough to include any kind of preparation, pre-processing and post-processing steps. We develop the guiding principles for the design of the ambitious framework and discuss its implementation. The QuaST decision tree includes multiple complete top-down paths from an application to its fully hybrid quantum solution.