Learning Quantum Processes with Memory -- Quantum Recurrent Neural Networks

Publikation: Arbeitspapier/PreprintPreprint

Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
PublikationsstatusElektronisch veröffentlicht (E-Pub) - 19 Jan. 2023

Abstract

Recurrent neural networks play an important role in both research and industry. With the advent of quantum machine learning, the quantisation of recurrent neural networks has become recently relevant. We propose fully quantum recurrent neural networks, based on dissipative quantum neural networks, capable of learning general causal quantum automata. A quantum training algorithm is proposed and classical simulations for the case of product outputs with the fidelity as cost function are carried out. We thereby demonstrate the potential of these algorithms to learn complex quantum processes with memory in terms of the exemplary delay channel, the time evolution of quantum states governed by a time-dependent Hamiltonian, and high- and low-frequency noise mitigation. Numerical simulations indicate that our quantum recurrent neural networks exhibit a striking ability to generalise from small training sets.

Zitieren

Learning Quantum Processes with Memory -- Quantum Recurrent Neural Networks. / Bondarenko, Dmytro; Salzmann, Robert; Schmiesing, Viktoria-S.
2023.

Publikation: Arbeitspapier/PreprintPreprint

Bondarenko D, Salzmann R, Schmiesing VS. Learning Quantum Processes with Memory -- Quantum Recurrent Neural Networks. 2023 Jan 19. Epub 2023 Jan 19. doi: 10.48550/ARXIV.2301.08167
Download
@techreport{ce2b0051407e42fd8118c19cecc232fe,
title = "Learning Quantum Processes with Memory -- Quantum Recurrent Neural Networks",
abstract = "Recurrent neural networks play an important role in both research and industry. With the advent of quantum machine learning, the quantisation of recurrent neural networks has become recently relevant. We propose fully quantum recurrent neural networks, based on dissipative quantum neural networks, capable of learning general causal quantum automata. A quantum training algorithm is proposed and classical simulations for the case of product outputs with the fidelity as cost function are carried out. We thereby demonstrate the potential of these algorithms to learn complex quantum processes with memory in terms of the exemplary delay channel, the time evolution of quantum states governed by a time-dependent Hamiltonian, and high- and low-frequency noise mitigation. Numerical simulations indicate that our quantum recurrent neural networks exhibit a striking ability to generalise from small training sets.",
author = "Dmytro Bondarenko and Robert Salzmann and Viktoria-S Schmiesing",
year = "2023",
month = jan,
day = "19",
doi = "10.48550/ARXIV.2301.08167",
language = "English",
type = "WorkingPaper",

}

Download

TY - UNPB

T1 - Learning Quantum Processes with Memory -- Quantum Recurrent Neural Networks

AU - Bondarenko, Dmytro

AU - Salzmann, Robert

AU - Schmiesing, Viktoria-S

PY - 2023/1/19

Y1 - 2023/1/19

N2 - Recurrent neural networks play an important role in both research and industry. With the advent of quantum machine learning, the quantisation of recurrent neural networks has become recently relevant. We propose fully quantum recurrent neural networks, based on dissipative quantum neural networks, capable of learning general causal quantum automata. A quantum training algorithm is proposed and classical simulations for the case of product outputs with the fidelity as cost function are carried out. We thereby demonstrate the potential of these algorithms to learn complex quantum processes with memory in terms of the exemplary delay channel, the time evolution of quantum states governed by a time-dependent Hamiltonian, and high- and low-frequency noise mitigation. Numerical simulations indicate that our quantum recurrent neural networks exhibit a striking ability to generalise from small training sets.

AB - Recurrent neural networks play an important role in both research and industry. With the advent of quantum machine learning, the quantisation of recurrent neural networks has become recently relevant. We propose fully quantum recurrent neural networks, based on dissipative quantum neural networks, capable of learning general causal quantum automata. A quantum training algorithm is proposed and classical simulations for the case of product outputs with the fidelity as cost function are carried out. We thereby demonstrate the potential of these algorithms to learn complex quantum processes with memory in terms of the exemplary delay channel, the time evolution of quantum states governed by a time-dependent Hamiltonian, and high- and low-frequency noise mitigation. Numerical simulations indicate that our quantum recurrent neural networks exhibit a striking ability to generalise from small training sets.

U2 - 10.48550/ARXIV.2301.08167

DO - 10.48550/ARXIV.2301.08167

M3 - Preprint

BT - Learning Quantum Processes with Memory -- Quantum Recurrent Neural Networks

ER -