A reinforcement learning approach for quantum state engineering
- PDF / 1,548,361 Bytes
- 14 Pages / 595.224 x 790.955 pts Page_size
- 0 Downloads / 235 Views
RESEARCH ARTICLE
A reinforcement learning approach for quantum state engineering ¨ Wrachtrup1,2 Jelena Mackeprang1 · Durga B. Rao Dasari1 · Jorg Received: 17 August 2019 / Accepted: 16 March 2020 © Springer Nature Switzerland AG 2020
Abstract Machine learning (ML) has become an attractive tool for solving various problems in different fields of physics, including the quantum domain. Here, we show how classical reinforcement learning (RL) could be used as a tool for quantum state engineering (QSE). We employ a measurement based control for QSE where the action sequences are determined by the choice of the measurement basis and the reward through the fidelity of obtaining the target state. Our analysis clearly displays a learning feature in QSE, for example in preparing arbitrary two-qubit entangled states and delivers successful action sequences that generalise previously found human solutions from exact quantum dynamics. We provide a systematic algorithmic approach for using RL for quantum protocols that deal with a non-trivial continuous state space, and discuss the scaling of these approaches for the preparation of larger entangled (cluster) states. Keywords Quantum state engineering · Quantum control · Deep reinforcement learning
1 Introduction Current advancements in quantum technology have rendered the research field and its applications increasingly complex. As machine learning (ML) has been successfully applied to various classical problems such as image recognition (Real et al. 2017; Krizhevsky et al. 2012), natural language processing (Lample et al. 2018) and board games (Silver et al. 2018), its applicability to quantum problems has gained increasing interest. Successful implementations of ML for problems in the field of quantum physics include neural networks for quantum state tomography (Torlai et al. 2018), detecting the speed-up of quantum walks (Melnikov et al. 2019) and classifying non-locality in quantum networks (Kriv´achy et al. 2019). There are three main variants of ML: supervised, unsupervised and reinforcement learning (RL). The latter is based on an abstract agent interacting with its environment Jelena Mackeprang
[email protected] Durga B. Rao Dasari [email protected] 1
3 Physikalisches Institut, Universit¨at Stuttgart, 70569 Stuttgart, Germany
2
Max Planck Institute for Solid State Research, 70569, Stuttgart, Germany
and receiving information in the form of a reward. It has already been used in the optimisation of quantum control (Bukov et al. 2018; Niu et al. 2019; Chen et al. 2014), quantum state preparation (Bukov 2018), coherent quantum state transport (Porotti et al. 2019), quantum error correction (Nautrup et al. 2018; Sweke et al. 2018; Andreasson et al. 2019; F¨osel et al. 2018) and much more. Entangled quantum states are crucial resources for quantum optics and information tasks such as precision sensing, quantum communication and computing. Construction and maintenance of such states have been a long-term challenge involving the clever use of interact
Data Loading...
