Andrew Barto

Andrew Barto is an American computer scientist whose work helped build the conceptual foundations of reinforcement learning. He studied mathematics at the University of Michigan and earned his doctorate in computer science there in 1975. His interests developed around using computational models to investigate learning and adaptive behavior.

Barto joined the University of Massachusetts Amherst in 1977 and developed a long academic career there. With colleagues and students, including Richard Sutton, he studied systems that learn from interaction rather than receiving an explicit correct answer for every example. This work connected ideas from psychology and neuroscience with the design of computational learning algorithms.

His research contributed to actor-critic methods, in which one part of a system chooses actions while another evaluates their likely consequences. Barto and Sutton also co-authored Reinforcement Learning: An Introduction. The textbook helped organize a field concerned with decisions whose effects unfold across a sequence of interactions rather than a single prediction.

Barto is professor emeritus at UMass Amherst and has received several major research honors. He shared the 2024 ACM A. M. Turing Award with Sutton, announced in 2025. His legacy includes algorithms, a widely used educational reference, and the training of researchers who extended reinforcement learning into robotics, games, and other forms of sequential decision-making.