How do real biological systems such as humans and other animals make decisions? This problem is the main focus of my research. In order to make decisions, the brain needs to make predictions about the consequences of the possible actions available to it. These predictions usually relate to the future expected reward (food, money etc.) that will be obtained from performing a particular action. The brain must then weigh up the different available actions and choose the action that gives the highest overall reward value. This is a complex computational problem which humans and other animals have evolved to solve in a near optimal manner. The goal of my research is to unravel the complex computations underlying reward prediction and decision making in the human brain. To aid in this process we make use of a variety of computational algorithms, including reinforcement learning, a family of models derived from computer science and robotics that seem to have strong relevance to how real animals make decisions. We combine these models with neural activity measured using fMRI (functional magnetic resonance imaging) so that we can reveal the computations in the brains of real human subjects while they perform decision making tasks.
Last updated: July 28, 2009 16:50
