Adiabatic quantum learning

TL;DR

This paper introduces adiabatic quantum learning, a novel approach that leverages adiabatic quantum evolution for quantum learning protocols, potentially enabling more efficient measurement processes in quantum computation.

Contribution

It proposes a new quantum learning framework based on adiabatic evolution, extending previous algorithms and suggesting integration with adiabatic weak measurement techniques.

Findings

Potential for single-measurement expectation value extraction

Extension of quantum learning algorithms to adiabatic processes

Illustrated with simple example protocols

Abstract

Adiabatic quantum control protocols have been of wide interest to quantum computation due to their robustness and insensitivity to their actual duration of execution. As an extension of previous quantum learning algorithms, this work proposes to execute some quantum learning protocols based entirely on adiabatic quantum evolution, hence dubbed as ``adiabatic quantum learning". In a conventional quantum machine learning protocol, the output is usually the expectation value of a pre-selected observable and the projective measurement of which forces a quantum circuit to run many times to obtain the output with a reasonable precision. By contrast, the proposed adiabatic quantum learning here may be integrated with future adiabatic weak measurement protocols, where a single measurement of the system allows to extract the expectation value of observables of interest without disrupting the concerned quantum states. Our main idea is illustrated with simple examples.