Computer Science Professor Andrew Barto is the recipient
of the 2004 IEEE Neural Networks Society Neural Networks Pioneer Award.
He was selected to receive this award for fundamental work on reinforcement
learning. The award was presented in July during the Opening Reception
of the 2004 International Joint Conference on Neural Networks in Budapest,
Barto’s research interests center on learning in both machines and animals. He has been developing learning algorithms that are useful for engineering applications while overlapping with learning as studied by experimental psychologists and neuroscientists. He has had a long interest in artificial and real neural networks. Most recently he has been working on three projects: the first focuses on extending reinforcement learning methods so that agents can autonomously construct hierarchies of reusable skills; the second, being conducted in collaboration with neuroscientists and developmental psychologists, involves modeling how animals learn motor skills; and the third applies machine learning methods to intelligent tutoring systems.