A simple model for learning in volatile environments

Payam Piray, Nathaniel D. Daw

Abstract: Author summary Sound principles of statistical learning dictate that uncertainty influences behavior. However, despite the success of statistically founded algorithms for learning in stable environments, in which uncertainty behaves in simple and predictable ways, it is challenging to develop a simple yet efficient algorithm for learning in volatile environments, in which uncertainty dynamically changes over time. In this article, we develop a model for learning in volatile environments. The proposed model is consistent with key concepts of classical learning theories from behavioral psychology. Furthermore, our model is algorithmically simpler, theoretically more accurate, and empirically more parsimonious than the state-of-the-art models of learning in volatile environments. The proposed model provides a coherent theory of learning under uncertainty.