Simple trees in complex forests: Growing Take The Best by Approximate Bayesian Computation

Advanced search
Browse by:

Department | Year

UCL Theses | Latest

Deposit your research

Simple trees in complex forests: Growing Take The Best by Approximate Bayesian Computation

Schulz, E; Speekenbrink, M; Meder, B; (2016) Simple trees in complex forests: Growing Take The Best by Approximate Bayesian Computation. In: Papafragou, A and Grodner, D and Mirman, D and Trueswell, JC, (eds.) Proceedings of the 38th Annual Meeting of the Cognitive Science Society 2016. (pp. pp. 2531-2536). Cognitive Science Society: Austin, TX. Green open access

Text
paper0437.pdf - Published Version
Available under License : See the attached licence file.
Download (489kB)

Abstract

How can heuristic strategies emerge from smaller building blocks? We propose Approximate Bayesian Computation (ABC) as a computational solution to this problem. As a first proof of concept, we demonstrate how a heuristic decision strategy such as Take The Best (TTB) can be learned from smaller, probabilistically updated building blocks. Based on a self-reinforcing sampling scheme, different building blocks are combined and, over time, tree-like non-compensatory heuristicsemerge. This new algorithm, coined Approximately Bayesian Computed Take The Best (ABC-TTB), is able to recover data that was generated by TTB, leads to sensible inferences about cue importance and cue directions, can outperform traditional TTB, and allows to trade-off performance and computational effort explicitly.

Type:	Proceedings paper
Title:	Simple trees in complex forests: Growing Take The Best by Approximate Bayesian Computation
Event:	38th Annual Meeting of the Cognitive Science Society 2016
ISBN-13:	9780991196739
Open access status:	An open access version is available from UCL Discovery
Publisher version:	https://mindmodeling.org/cogsci2016/papers/0437/pa...
Language:	English
Additional information:	Copyright © The Authors 2016
Keywords:	Heuristics, Take The Best, Approximate Bayesian Computation, Reinforcement Learning
UCL classification:	UCL UCL > Provost and Vice Provost Offices 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 > Div of Psychology and Lang Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Brain Sciences > Div of Psychology and Lang Sciences > Experimental Psychology
URI:	https://discovery.ucl.ac.uk/id/eprint/1512470