Neural network models of decision making with learning on multiple timescales

Animals can quickly learn to make appropriate decisions according to their environment that can change over a wide range of timescales. Yet the neural computation underling the adaptive decision making is not well understood. To investigate basic computational principles and neural mechanisms, here...

Full description

Bibliographic Details
Main Author: Iigaya, Kiyohito
Language:English
Published: 2014
Subjects:
Online Access:https://doi.org/10.7916/D8BR8Q81
Description
Summary:Animals can quickly learn to make appropriate decisions according to their environment that can change over a wide range of timescales. Yet the neural computation underling the adaptive decision making is not well understood. To investigate basic computational principles and neural mechanisms, here we study simple neural network models for decision making with learning on multiple timescales, and we test our model's predictions in experimental data. We provide basic network models for value-based decision making under uncertainty.