Proceedings of the Annual Meeting of the Cognitive Science Society

We explore a nonparametric approach to cognitive modeling.Traditionally, models in cognitive science have been paramet-ric. As such, the model relies on the assumption that the datadistribution can be defined by a finite set of parameters. How-ever, there is no guarantee that such an assumption will hold,and it may introduce undesirable biases. For these reasons, anonparametric approach to model building is appealing. Wepropose a novel framework that combines Gaussian Processeswith active learning (GPAL), and evaluate it in the context ofdelay discounting (DD), a well-studied task in decision mak-ing. We evaluate GPAL in a simulation and a behavioral exper-iment, and compare it against a traditional parametric model.The results show that GPAL is a suitable modeling frameworkthat is robust, reliable, and efficient, exhibiting high sensitivityto individual differences.