English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  Two steps to risk sensitivity

Gagne, C., & Dayan, P. (in press). Two steps to risk sensitivity. In M. Ranzato, A. Beygelzimer, P. Liang, J. Vaughan, & Y. Dauphin (Eds.), Advances in Neural Information Processing Systems 34.

Item is

Basic

show hide
Genre: Conference Paper

Files

show Files

Locators

show
hide
Description:
-

Creators

show
hide
 Creators:
Gagne, C1, 2, Author              
Dayan, P1, 2, Author              
Affiliations:
1Department of Computational Neuroscience, Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_3017468              
2Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_1497794              

Content

show
hide
Free keywords: -
 Abstract: Distributional reinforcement learning (RL) – in which agents learn about all the possible long-term consequences of their actions, and not just the expected value – is of great recent interest. One of the most important affordances of a distributional view is facilitating a modern, measured, approach to risk when outcomes are not completely certain. By contrast, psychological and neuroscientific investigations into decision making under risk have utilized a variety of more venerable theoretical models such as prospect theory that lack axiomatically desirable properties such as coherence. Here, we consider a particularly relevant risk measure for modeling human and animal planning, called conditional value-at-risk (CVaR), which quantifies worst-case outcomes (e.g., vehicle accidents or predation). We first adopt a conventional distributional approach to CVaR in a sequential setting and reanalyze the choices of human decision-makers in the well-known two-step task, revealing substantial risk aversion that had been lurking under stickiness and perseveration. We then consider a further critical property of risk sensitivity, namely time consistency, showing alternatives to this form of CVaR that enjoy this desirable characteristic. We use simulations to examine settings in which the various forms differ in ways that have implications for human and animal planning and behavior.

Details

show
hide
Language(s):
 Dates: 2021-10
 Publication Status: Accepted / In Press
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: -
 Degree: -

Event

show
hide
Title: Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021)
Place of Event: -
Start-/End Date: 2021-12-06 - 2021-12-14

Legal Case

show

Project information

show

Source 1

show
hide
Title: Advances in Neural Information Processing Systems 34
Source Genre: Proceedings
 Creator(s):
Ranzato, M, Editor
Beygelzimer, A, Editor
Liang, PS, Editor
Vaughan, JW, Editor
Dauphin, Y, Editor
Affiliations:
-
Publ. Info: -
Pages: - Volume / Issue: - Sequence Number: - Start / End Page: - Identifier: -