Ultrafast photonic reinforcement learning based on laser chaos

Abstract Reinforcement learning involves decision making in dynamic and uncertain environments and constitutes an important element of artificial intelligence (AI). In this work, we experimentally demonstrate that the ultrafast chaotic oscillatory dynamics of lasers efficiently solve the multi-armed...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Makoto Naruse, Yuta Terashima, Atsushi Uchida, Song-Ju Kim
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2017
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/506a6a3704f44445ad9f4c41e8098dc9
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:506a6a3704f44445ad9f4c41e8098dc9
record_format	dspace
spelling	oai:doaj.org-article:506a6a3704f44445ad9f4c41e8098dc92021-12-02T16:08:13ZUltrafast photonic reinforcement learning based on laser chaos10.1038/s41598-017-08585-82045-2322https://doaj.org/article/506a6a3704f44445ad9f4c41e8098dc92017-08-01T00:00:00Zhttps://doi.org/10.1038/s41598-017-08585-8https://doaj.org/toc/2045-2322Abstract Reinforcement learning involves decision making in dynamic and uncertain environments and constitutes an important element of artificial intelligence (AI). In this work, we experimentally demonstrate that the ultrafast chaotic oscillatory dynamics of lasers efficiently solve the multi-armed bandit problem (MAB), which requires decision making concerning a class of difficult trade-offs called the exploration–exploitation dilemma. To solve the MAB, a certain degree of randomness is required for exploration purposes. However, pseudorandom numbers generated using conventional electronic circuitry encounter severe limitations in terms of their data rate and the quality of randomness due to their algorithmic foundations. We generate laser chaos signals using a semiconductor laser sampled at a maximum rate of 100 GSample/s, and combine it with a simple decision-making principle called tug of war with a variable threshold, to ensure ultrafast, adaptive, and accurate decision making at a maximum adaptation speed of 1 GHz. We found that decision-making performance was maximized with an optimal sampling interval, and we highlight the exact coincidence between the negative autocorrelation inherent in laser chaos and decision-making performance. This study paves the way for a new realm of ultrafast photonics in the age of AI, where the ultrahigh bandwidth of light wave can provide new value.Makoto NaruseYuta TerashimaAtsushi UchidaSong-Ju KimNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 7, Iss 1, Pp 1-10 (2017)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Makoto Naruse Yuta Terashima Atsushi Uchida Song-Ju Kim Ultrafast photonic reinforcement learning based on laser chaos
description	Abstract Reinforcement learning involves decision making in dynamic and uncertain environments and constitutes an important element of artificial intelligence (AI). In this work, we experimentally demonstrate that the ultrafast chaotic oscillatory dynamics of lasers efficiently solve the multi-armed bandit problem (MAB), which requires decision making concerning a class of difficult trade-offs called the exploration–exploitation dilemma. To solve the MAB, a certain degree of randomness is required for exploration purposes. However, pseudorandom numbers generated using conventional electronic circuitry encounter severe limitations in terms of their data rate and the quality of randomness due to their algorithmic foundations. We generate laser chaos signals using a semiconductor laser sampled at a maximum rate of 100 GSample/s, and combine it with a simple decision-making principle called tug of war with a variable threshold, to ensure ultrafast, adaptive, and accurate decision making at a maximum adaptation speed of 1 GHz. We found that decision-making performance was maximized with an optimal sampling interval, and we highlight the exact coincidence between the negative autocorrelation inherent in laser chaos and decision-making performance. This study paves the way for a new realm of ultrafast photonics in the age of AI, where the ultrahigh bandwidth of light wave can provide new value.
format	article
author	Makoto Naruse Yuta Terashima Atsushi Uchida Song-Ju Kim
author_facet	Makoto Naruse Yuta Terashima Atsushi Uchida Song-Ju Kim
author_sort	Makoto Naruse
title	Ultrafast photonic reinforcement learning based on laser chaos
title_short	Ultrafast photonic reinforcement learning based on laser chaos
title_full	Ultrafast photonic reinforcement learning based on laser chaos
title_fullStr	Ultrafast photonic reinforcement learning based on laser chaos
title_full_unstemmed	Ultrafast photonic reinforcement learning based on laser chaos
title_sort	ultrafast photonic reinforcement learning based on laser chaos
publisher	Nature Portfolio
publishDate	2017
url	https://doaj.org/article/506a6a3704f44445ad9f4c41e8098dc9
work_keys_str_mv	AT makotonaruse ultrafastphotonicreinforcementlearningbasedonlaserchaos AT yutaterashima ultrafastphotonicreinforcementlearningbasedonlaserchaos AT atsushiuchida ultrafastphotonicreinforcementlearningbasedonlaserchaos AT songjukim ultrafastphotonicreinforcementlearningbasedonlaserchaos
_version_	1718384593176887296

Ultrafast photonic reinforcement learning based on laser chaos

Ejemplares similares