Associative Learning Should Go Deep

Esther Mondragón; Eduardo Alonso; Niklas Kokkola

doi:10.1016/j.tics.2017.06.001

Associative Learning Should Go Deep

Trends Cogn Sci. 2017 Nov;21(11):822-825. doi: 10.1016/j.tics.2017.06.001. Epub 2017 Jun 28.

Authors

Esther Mondragón¹, Eduardo Alonso², Niklas Kokkola³

Affiliations

¹ Research Centre for Systems and Control, University of London, London EC1V 0HB, UK; Computational and Animal Learning Research Centre, St Albans AL1 1RQ, UK. Electronic address: e.mondragon@cal-r.org.
² Research Centre for Systems and Control, University of London, London EC1V 0HB, UK; Computational and Animal Learning Research Centre, St Albans AL1 1RQ, UK. Electronic address: E.Alonso@city.ac.uk.
³ Research Centre for Systems and Control, University of London, London EC1V 0HB, UK; Computational and Animal Learning Research Centre, St Albans AL1 1RQ, UK. Electronic address: Niklas.Kokkola.1@city.ac.uk.

PMID: 28668210
DOI: 10.1016/j.tics.2017.06.001

Abstract

Conditioning, how animals learn to associate two or more events, is one of the most influential paradigms in learning theory. It is nevertheless unclear how current models of associative learning can accommodate complex phenomena without ad hoc representational assumptions. We propose to embrace deep neural networks to negotiate this problem.

Keywords: associative learning; deep neural networks.

MeSH terms

Animals
Association Learning*
Humans
Neural Networks, Computer*