Performing Bayesian inference with exemplar models

Probabilistic models have recently received much attention as accounts of human cognition. However, previous work has fo-cused on formulating the abstract problems behind cognitive tasks and their probabilistic solutions, rather than considering mechanisms that could implement these solutions. Exemplar models are a successful class of psychological process mod-els that use an inventory of stored examples to solve prob-lems such as identification, categorization and function learn-ing. We show that exemplar models can be interpreted as a sophisticated form of Monte Carlo approximation known as importance sampling, and thus provide a way to perform ap-proximate Bayesian inference. Simulations of Bayesian infer-ence in speech perception and concept learning show that ex-emplar models can account for human performance with only a few exemplars, for both simple and relatively complex prior distributions. Thus, we show that exemplar models provide a possible mechanism for implementing Bayesian inference