The Respective Roles of Priors, Noise, and Confidence in Learning
Abstract
Learning and deciding in uncertain environments is a difficult but ubiquitous challenge for human intelligence. Research in psychology and neuroscience has identified the elementary cognitive processes (called inferences) that we use every day in a multitude of varied situations to integrate ambiguous information and act appropriately. These inferences are surprisingly imprecise (noisy in statistical terms), which places severe constraints on the accuracy and predictability of human decisions made in uncertain environments. During this seminar, drawing on recent findings in psychology and neuroscience, I will present the respective roles of a priori, inference noise and confidence in human learning and decision-making under uncertainty.
