Machine learning and its applications in quantum physics: metrology, thermology, phase transitions and dynamical decoupling

Abstract

This research project aims to develop a methodology that can be applied to four main themes of quantum physics: quantum metrology, quantum thermology, error protection techniques and quantum phase transitions. We will focus on several machine learning techniques, considering from the simplest to the most sophisticated models. Concerning quantum metrology, we will focus on the study of quantum dots, trying to develop an algorithm capable of determining the constant coupling between two qubits. In thermology, our objective will be to determine the temperature of a reservoir by means of an auxiliar system. In the case of phase transitions, our focus will be on the development of a universal classifier, that is, an algorithm capable of detecting the phases even of unknown models. Finally, we will make use of machine learning to optimize quantum information protection techniques, in particular dynamical decoupling. During the development of this project, we intend to introduce new routines for the study of metrology, thermology, quantum phase transitions and protection techniques guided now by computational methods based on machine learning. (AU)