Invited Talks

Models for sequential decision making under uncertainty (such as Markov decision processes, or MDPs) have been studied in operations research for decades. The recent incorporation of ideas from many areas of AI, including planning, probabilistic modeling, machine learning, and knowledge representation, have made these models much more widely applicable. In this talk, Boutilier will survey recent advances within AI in the use of fullyand partially-observable MDPs as a modeling tool, and the development of computationallymanageable solution methods. He will place special emphasis on algorithms that exploit specific problem structure and approximation techniques.