License: Creative Commons Attribution 4.0 International license (CC BY 4.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.TQC.2023.13
URN: urn:nbn:de:0030-drops-183230
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2023/18323/
Go to the corresponding LIPIcs Volume Portal

Jerbi, Sofiene ; Cornelissen, Arjan ; Ozols, Maris ; Dunjko, Vedran

Quantum Policy Gradient Algorithms

pdf-format:
LIPIcs-TQC-2023-13.pdf (1.0 MB)

Abstract

Understanding the power and limitations of quantum access to data in machine learning tasks is primordial to assess the potential of quantum computing in artificial intelligence. Previous works have already shown that speed-ups in learning are possible when given quantum access to reinforcement learning environments. Yet, the applicability of quantum algorithms in this setting remains very limited, notably in environments with large state and action spaces. In this work, we design quantum algorithms to train state-of-the-art reinforcement learning policies by exploiting quantum interactions with an environment. However, these algorithms only offer full quadratic speed-ups in sample complexity over their classical analogs when the trained policies satisfy some regularity conditions. Interestingly, we find that reinforcement learning policies derived from parametrized quantum circuits are well-behaved with respect to these conditions, which showcases the benefit of a fully-quantum reinforcement learning framework.

BibTeX - Entry

@InProceedings{jerbi_et_al:LIPIcs.TQC.2023.13,
  author =	{Jerbi, Sofiene and Cornelissen, Arjan and Ozols, Maris and Dunjko, Vedran},
  title =	{{Quantum Policy Gradient Algorithms}},
  booktitle =	{18th Conference on the Theory of Quantum Computation, Communication and Cryptography (TQC 2023)},
  pages =	{13:1--13:24},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-283-9},
  ISSN =	{1868-8969},
  year =	{2023},
  volume =	{266},
  editor =	{Fawzi, Omar and Walter, Michael},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/opus/volltexte/2023/18323},
  URN =		{urn:nbn:de:0030-drops-183230},
  doi =		{10.4230/LIPIcs.TQC.2023.13},
  annote =	{Keywords: quantum reinforcement learning, policy gradient methods, parametrized quantum circuits}
}

Keywords: quantum reinforcement learning, policy gradient methods, parametrized quantum circuits
Collection: 18th Conference on the Theory of Quantum Computation, Communication and Cryptography (TQC 2023)
Issue Date: 2023
Date of publication: 18.07.2023

DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI