Computational Models of Decision-Making

We make decisions countless times a day. Computational models have been developed that improve our understanding of how these decisions are made. This course is organized in three parts: perceptual decision-making, value-based decision-making, and computational psychiatry. In part one, perceptual decision-making, we will focus on computational models that can capture and explain decisions in perception, such as categorizing an orientation, or discriminating the direction of moving dots, or estimating the magnitude of a stimulus (e.g., time). We will start by laying the foundations of signal detection theory and Bayesian inference under uncertainty and build to models that incorporate confidence ratings and reaction times. In part two, value-based decision-making, we will move on to decisions that incorporate our values (e.g., ‘Should I go out or stay in and study?’, ‘Should I eat a burger or a salad?’). We will learn the basics of a computational modeling framework that captures how we learn values from rewards and punishments, reinforcement learning, as well as about model-free and model-based learning. Lastly, we will learn how impairments in decision-making that occur in psychopathology (e.g., addiction, anorexia nervosa, anxiety) have been conceptualized and quantified in the relatively new field of computational psychiatry.