Learning from imperfect quantum data via unsupervised domain adaptation with classical shadows

2026-03-30

AI Summary

This work addresses the performance degradation of quantum machine learning models in real-world scenarios, where target-domain quantum data often lack clean and complete labels, leading to a mismatch between training and deployment conditions. To tackle this challenge, the study introduces unsupervised domain adaptation into quantum machine learning for the first time. By leveraging classical shadows to map quantum states into classical representations, the method enables effective learning from imperfect target-domain data entirely within a classical computational pipeline. This approach relaxes the conventional reliance on ideally labeled data and demonstrates significant improvements over non-adaptive baselines trained solely on source-domain data as well as unsupervised methods using only target-domain data, as validated on tasks of quantum phase recognition and entanglement classification.

Abstract

Learning from quantum data using classical machine learning models has emerged as a promising paradigm toward realizing quantum advantages. Despite extensive analyses on their performance, clean and fully labeled quantum data from the target domain are often unavailable in practical scenarios, forcing models to be trained on data collected under conditions that differ from those encountered at deployment. This mismatch highlights the need for new approaches beyond the common assumptions of prior work. In this work, we address this issue by employing an unsupervised domain adaptation framework for learning from imperfect quantum data. Specifically, by leveraging classical representations of quantum states obtained via classical shadows, we perform unsupervised domain adaptation entirely within a classical computational pipeline once measurements on the quantum states are executed. We numerically evaluate the framework on quantum phases of matter and entanglement classification tasks under realistic domain shifts. Across both tasks, our method outperforms source-only non-adaptive baselines and target-only unsupervised learning approaches, demonstrating the practical applicability of domain adaptation to realistic quantum data learning.

Problem

• quantum data
• domain adaptation
• imperfect data
• unsupervised learning
• distribution mismatch

Innovation

• unsupervised domain adaptation
• classical shadows
• quantum data learning
• domain shift
• classical machine learning