People are often faced with choices where there is a conflictbetween seeking reward and gathering information. In manyof these cases there exists a functional relationship betweenthe features associated with actions and their correspondingrewards. Accounts of how people make decisions in thesecircumstances have not considered how peoples’ strategiesdepend on the complexity of this function, as well as theperson’s goal. In a sequential decision making task we foundthat people chose between a number of different explorationstrategies, but that strategy selection did not necessarily alignwith goal or account for function complexity.

Whether learning how pressing on the gas pedal of rental car will affect its acceleration or learning how changingthe volume of speakers affects the perceived loudness of the sound they produce, humans can quickly learn functions from afew examples. Recent hybrid models (Lucas et al., 2015) combine the structure of rule-based models with the flexibility ofsimilarity-based models by exploiting the equivalence of Bayesian linear regression and Gaussian processes. We expand onthese models by taking advantage of the compositional nature of Gaussian processes and imposing a generative grammar overa set of base components in order to build the structured but diverse hypothesis spaces that appear to be represented by people.Subsequent testing will compare this model’s ability to reproduce people’s learning difficulty rankings of different functions,extrapolation results, and representations of multiple overlapping functions to that of other hybrid models.