A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise

Abstract Bayesian models have advanced the idea that humans combine prior beliefs and sensory observations to optimize behavior. How the brain implements Bayes-optimal inference, however, remains poorly understood. Simple behavioral tasks suggest that the brain can flexibly represent probability dis...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Seth W. Egger, Mehrdad Jazayeri
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2018
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/f163ec5b6de640b5b546097fecc19846
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:f163ec5b6de640b5b546097fecc19846
record_format	dspace
spelling	oai:doaj.org-article:f163ec5b6de640b5b546097fecc198462021-12-02T11:40:26ZA nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise10.1038/s41598-018-30722-02045-2322https://doaj.org/article/f163ec5b6de640b5b546097fecc198462018-08-01T00:00:00Zhttps://doi.org/10.1038/s41598-018-30722-0https://doaj.org/toc/2045-2322Abstract Bayesian models have advanced the idea that humans combine prior beliefs and sensory observations to optimize behavior. How the brain implements Bayes-optimal inference, however, remains poorly understood. Simple behavioral tasks suggest that the brain can flexibly represent probability distributions. An alternative view is that the brain relies on simple algorithms that can implement Bayes-optimal behavior only when the computational demands are low. To distinguish between these alternatives, we devised a task in which Bayes-optimal performance could not be matched by simple algorithms. We asked subjects to estimate and reproduce a time interval by combining prior information with one or two sequential measurements. In the domain of time, measurement noise increases with duration. This property takes the integration of multiple measurements beyond the reach of simple algorithms. We found that subjects were able to update their estimates using the second measurement but their performance was suboptimal, suggesting that they were unable to update full probability distributions. Instead, subjects’ behavior was consistent with an algorithm that predicts upcoming sensory signals, and applies a nonlinear function to errors in prediction to update estimates. These results indicate that the inference strategies employed by humans may deviate from Bayes-optimal integration when the computational demands are high.Seth W. EggerMehrdad JazayeriNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 8, Iss 1, Pp 1-16 (2018)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Seth W. Egger Mehrdad Jazayeri A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
description	Abstract Bayesian models have advanced the idea that humans combine prior beliefs and sensory observations to optimize behavior. How the brain implements Bayes-optimal inference, however, remains poorly understood. Simple behavioral tasks suggest that the brain can flexibly represent probability distributions. An alternative view is that the brain relies on simple algorithms that can implement Bayes-optimal behavior only when the computational demands are low. To distinguish between these alternatives, we devised a task in which Bayes-optimal performance could not be matched by simple algorithms. We asked subjects to estimate and reproduce a time interval by combining prior information with one or two sequential measurements. In the domain of time, measurement noise increases with duration. This property takes the integration of multiple measurements beyond the reach of simple algorithms. We found that subjects were able to update their estimates using the second measurement but their performance was suboptimal, suggesting that they were unable to update full probability distributions. Instead, subjects’ behavior was consistent with an algorithm that predicts upcoming sensory signals, and applies a nonlinear function to errors in prediction to update estimates. These results indicate that the inference strategies employed by humans may deviate from Bayes-optimal integration when the computational demands are high.
format	article
author	Seth W. Egger Mehrdad Jazayeri
author_facet	Seth W. Egger Mehrdad Jazayeri
author_sort	Seth W. Egger
title	A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
title_short	A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
title_full	A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
title_fullStr	A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
title_full_unstemmed	A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
title_sort	nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise
publisher	Nature Portfolio
publishDate	2018
url	https://doaj.org/article/f163ec5b6de640b5b546097fecc19846
work_keys_str_mv	AT sethwegger anonlinearupdatingalgorithmcapturessuboptimalinferenceinthepresenceofsignaldependentnoise AT mehrdadjazayeri anonlinearupdatingalgorithmcapturessuboptimalinferenceinthepresenceofsignaldependentnoise AT sethwegger nonlinearupdatingalgorithmcapturessuboptimalinferenceinthepresenceofsignaldependentnoise AT mehrdadjazayeri nonlinearupdatingalgorithmcapturessuboptimalinferenceinthepresenceofsignaldependentnoise
_version_	1718395617198211072

A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise

Ejemplares similares