Matthew Botvinick, Princeton

 

Matthew Botvinick, Princeton

Sage 4101

December 2, 2009 12:00 PM - 1:30 PM


Abstract:  Recent work has given rise to the view that reward-based decision making is governed by two key controllers: a habit system, which stores stimulus-response associations shaped by past reward, and a goal-oriented system that selects actions based on their anticipated outcomes. The current literature provides a rich body of computational theory addressing habit formation, centering on temporal-difference learning mechanisms. Less progress has been made toward formalizing the processes involved in goal-directed decision making. We draw on recent work in cognitive neuroscience, animal learning, cognitive and developmental psychology and machine learning, to propose a new theory of goal-directed decision making. Our basic proposal is that the brain, within an identifiable network of cortical and subcortical structures, implements a generative model of action: a representation of the interdependencies among states, actions, outcomes, and rewards. Goal-directed decision making is understood as involving a set of procedures for querying this generative model, in order to extract plans for action. We focus on one particularly interesting and powerful form of query, which involves reasoning from the assumption of future reward. The proposed account has been formalized in a set of simulations addressing benchmark empirical phenomena, and tested through two human behavioral studies. I'll discuss the relationship between the proposed framework and other models of decision making, including recent models of perceptual decision making, to which our theory bears an intimate but surprising connection.  

 
Add to calendar
Share|