Dissociation between asymmetric value updating and perseverance in human reinforcement learning

Abstract The learning rate is a key parameter in reinforcement learning that determines the extent to which novel information (outcome) is incorporated in guiding subsequent actions. Numerous studies have reported that the magnitude of the learning rate in human reinforcement learning is biased depe...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Michiyo Sugawara, Kentaro Katahira
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/bd93cbb47f574d3f9424c9e878b5b931
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:bd93cbb47f574d3f9424c9e878b5b931
record_format	dspace
spelling	oai:doaj.org-article:bd93cbb47f574d3f9424c9e878b5b9312021-12-02T12:09:45ZDissociation between asymmetric value updating and perseverance in human reinforcement learning10.1038/s41598-020-80593-72045-2322https://doaj.org/article/bd93cbb47f574d3f9424c9e878b5b9312021-02-01T00:00:00Zhttps://doi.org/10.1038/s41598-020-80593-7https://doaj.org/toc/2045-2322Abstract The learning rate is a key parameter in reinforcement learning that determines the extent to which novel information (outcome) is incorporated in guiding subsequent actions. Numerous studies have reported that the magnitude of the learning rate in human reinforcement learning is biased depending on the sign of the reward prediction error. However, this asymmetry can be observed as a statistical bias if the fitted model ignores the choice autocorrelation (perseverance), which is independent of the outcomes. Therefore, to investigate the genuine process underlying human choice behavior using empirical data, one should dissociate asymmetry in learning and perseverance from choice behavior. The present study addresses this issue by using a Hybrid model incorporating asymmetric learning rates and perseverance. First, by conducting simulations, we demonstrate that the Hybrid model can identify the true underlying process. Second, using the Hybrid model, we show that empirical data collected from a web-based experiment are governed by perseverance rather than asymmetric learning. Finally, we apply the Hybrid model to two open datasets in which asymmetric learning was reported. As a result, the asymmetric learning rate was validated in one dataset but not another.Michiyo SugawaraKentaro KatahiraNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-13 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Michiyo Sugawara Kentaro Katahira Dissociation between asymmetric value updating and perseverance in human reinforcement learning
description	Abstract The learning rate is a key parameter in reinforcement learning that determines the extent to which novel information (outcome) is incorporated in guiding subsequent actions. Numerous studies have reported that the magnitude of the learning rate in human reinforcement learning is biased depending on the sign of the reward prediction error. However, this asymmetry can be observed as a statistical bias if the fitted model ignores the choice autocorrelation (perseverance), which is independent of the outcomes. Therefore, to investigate the genuine process underlying human choice behavior using empirical data, one should dissociate asymmetry in learning and perseverance from choice behavior. The present study addresses this issue by using a Hybrid model incorporating asymmetric learning rates and perseverance. First, by conducting simulations, we demonstrate that the Hybrid model can identify the true underlying process. Second, using the Hybrid model, we show that empirical data collected from a web-based experiment are governed by perseverance rather than asymmetric learning. Finally, we apply the Hybrid model to two open datasets in which asymmetric learning was reported. As a result, the asymmetric learning rate was validated in one dataset but not another.
format	article
author	Michiyo Sugawara Kentaro Katahira
author_facet	Michiyo Sugawara Kentaro Katahira
author_sort	Michiyo Sugawara
title	Dissociation between asymmetric value updating and perseverance in human reinforcement learning
title_short	Dissociation between asymmetric value updating and perseverance in human reinforcement learning
title_full	Dissociation between asymmetric value updating and perseverance in human reinforcement learning
title_fullStr	Dissociation between asymmetric value updating and perseverance in human reinforcement learning
title_full_unstemmed	Dissociation between asymmetric value updating and perseverance in human reinforcement learning
title_sort	dissociation between asymmetric value updating and perseverance in human reinforcement learning
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/bd93cbb47f574d3f9424c9e878b5b931
work_keys_str_mv	AT michiyosugawara dissociationbetweenasymmetricvalueupdatingandperseveranceinhumanreinforcementlearning AT kentarokatahira dissociationbetweenasymmetricvalueupdatingandperseveranceinhumanreinforcementlearning
_version_	1718394654072766464

Dissociation between asymmetric value updating and perseverance in human reinforcement learning

Ejemplares similares