Kurth-Nelson, Z;
Redish, AD;
(2009)
Temporal-Difference Reinforcement Learning with Distributed Representations.
PLOS ONE
, 4
(10)
, Article e7362. 10.1371/journal.pone.0007362.
![[thumbnail of 1346634.pdf]](https://discovery.ucl.ac.uk/style/images/fileicons/application_pdf.png) Preview |
PDF
1346634.pdf Download (1MB) |
Abstract
Formal Correction: This article has been formally corrected to address the following errors. 1.The initials for author A. David Redish appear incorrectly in the Author Contributions. The correct initials are ADR. The Author Contributions should read: "Conceived and designed the experiments: ZKN DDR. Performed the experiments: ZKN ADR Temporal-difference (TD) algorithms have been proposed as models of reinforcement learning (RL). We examine two issues of distributed representation in these TD algorithms: distributed representations of belief and distributed discounting factors. Distributed representation of belief allows the believed state of the world to distribute across sets of equivalent states. Distributed exponential discounting factors produce hyperbolic discounting in the behavior of the agent itself. We examine these issues in the context of a TD RL model in which state-belief is distributed over a set of exponentially-discounting "micro-Agents", each of which has a separate discounting factor (gamma). Each mu Agent maintains an independent hypothesis about the state of the world, and a separate value-estimate of taking actions within that hypothesized state. The overall agent thus instantiates a flexible representation of an evolving world-state. As with other TD models, the value-error (delta) signal within the model matches dopamine signals recorded from animals in standard conditioning reward-paradigms. The distributed representation of belief provides an explanation for the decrease in dopamine at the conditioned stimulus seen in overtrained animals, for the differences between trace and delay conditioning, and for transient bursts of dopamine seen at movement initiation. Because each mAgent also includes its own exponential discounting factor, the overall agent shows hyperbolic discounting, consistent with behavioral experiments.
| Type: | Article |
|---|---|
| Title: | Temporal-Difference Reinforcement Learning with Distributed Representations |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1371/journal.pone.0007362 |
| Publisher version: | http://dx.doi.org/10.1371/journal.pone.0007362 |
| Language: | English |
| Additional information: | © 2009 Kurth-Nelson, Redish. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. This work was supported by a fellowship on NSF-IGERT #9870633 (ZKN), by a Sloan Fellowship (ADR), and by NIH DA024080, as well as by a Career Development Award from the University of Minnesota Transdisciplinary Tobacco Use Research Center (TTURC) (ADR). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. |
| Keywords: | MIDBRAIN DOPAMINE NEURONS, BASAL GANGLIA, NUCLEUS-ACCUMBENS, REWARD-PREDICTION, DECISION-MAKING, DELAYED REWARDS, NEURAL ACTIVITY, FUTURE REWARDS, HUMAN BRAIN, HIPPOCAMPUS |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Brain Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Brain Sciences > UCL Queen Square Institute of Neurology UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Brain Sciences > UCL Queen Square Institute of Neurology > Imaging Neuroscience |
| URI: | https://discovery.ucl.ac.uk/id/eprint/1346634 |
Archive Staff Only
 |
View Item |
