On the asymptotic equivalence between differential hebbian and temporal difference learning

Kolodziejski, C., Porr, B. and Worgotter, F. (2009) On the asymptotic equivalence between differential hebbian and temporal difference learning. Neural Computation, 21(4), pp. 1173-1202. (doi:10.1162/neco.2008.04-08-750)

Kolodziejski, C., Porr, B. and Worgotter, F. (2009) On the asymptotic equivalence between differential hebbian and temporal difference learning. Neural Computation, 21(4), pp. 1173-1202. (doi:10.1162/neco.2008.04-08-750)

Full text not currently available from Enlighten.

Abstract

In this theoretical contribution, we provide mathematical proof that two of the most important classes of network learning-correlation-based differential Hebbian learning and reward-based temporal difference learning-are asymptotically equivalent when timing the learning with a modulatory signal. This opens the opportunity to consistently reformulate most of the abstract reinforcement learning framework from a correlation-based perspective more closely related to the biophysics of neurons

Item Type:Articles
Keywords:Basal ganglia, expectation, midbrain dopamine, model, networks, neuronal-activity, reward, science, striatum, systems, timing-dependent plasticity
Status:Published
Refereed:Yes
Glasgow Author(s) Enlighten ID:Porr, Dr Bernd
Authors: Kolodziejski, C., Porr, B., and Worgotter, F.
College/School:College of Science and Engineering > School of Engineering > Biomedical Engineering
Journal Name:Neural Computation
Journal Abbr.:Neural comp.
ISSN:0899-7667
ISSN (Online):1530-888X
Published Online:20 March 2009

University Staff: Request a correction | Enlighten Editors: Update this record