AutoQML - a Framework for Automated Quantum Machine Learning

In this work, we present AutoQML, a framework that seamlessly integrates Quantum Machine Learning (QML) algorithms into Automated Machine Learning (AutoML). Leveraging the advantages of the AutoML paradigm, the framework is intentionally designed with a high level of abstraction, eliminating the need for users to possess extensive experience in both Machine Learning (ML) and Quantum Computing (QC). The tool automates the entire process of constructing typical ML pipelines including data cleaning and preprocessing as well as model selection, optimization, and evaluation. Additionally, it automatizes QC-specific aspects as for example selection of quantum backends and execution management on real quantum hardware. AutoQML utilizes Ray as its underlying AutoML optimization framework and employs the in-house developed QML library sQUlearn for providing QML algorithms. Both of these components provide low-level functionality and can be used as standalone solutions. Finally, we delve into the integration steps required to incorporate the framework into the Quantum Computing-as-a-Service platform, PlankQK.