Computational models of human learning, teaching, and trust

Abstract

Note the updated date/time/location for this talk

Humans are remarkable for their ability to learn so much so quickly. I will discuss my research, which investigates the power of human learning with a combination of computational models and behavioral experiments. I will focus on three core problems: flexibility in learning about the world, learning from knowledgeable and helpful informants, and joint learning about the world and informants. I will conclude by discussing implications of this research for problems of data analysis, learning from the non-random (intentional) actions, and for leveraging social reasoning toward educational goals.