摘要

We consider algorithms that aim to find a linear function approximation to the cost function of a policy in infinite-horizon Markovian Decision Problems. We discuss extensions/new versions of the standard temporal difference (TD) methods, which offer special implementation advantages and reduced overhead. We then discuss how these methods can be extended for approximate solution of high dimensional fixed point problems within low-dimensional subspaces.

演讲人简历

Dimitri P. Bertsekas received his undergraduate degree in engineering from the National Technical University of Athens, Greece, and his Ph.D. from the Massachusetts Institute of Technology.

Dr. Bertsekas has held faculty positions with the Engineering-Economic Systems Dept., Stanford University (1971-1974) and the Electrical Engineering Dept. of the University of Illinois, Urbana  (1974-1979). Since 1979 he has been teaching at the Electrical Engineering and Computer Science Department of the Massachusetts Institute of Technology (M.I.T.), where he is currently McAfee Professor of Engineering. He consults regularly with private industry and has held editorial positions in several journals. His research at  M.I.T. spans several fields, including optimization, control, large-scale computation, and data communication networks, and is closely  tied to his teaching and book authoring activities. He has written numerous research papers, and fourteen books, several of which are used as textbooks in MIT classes.

Professor Bertsekas was awarded the INFORMS 1997 Prize for Research  Excellence in the Interface Between Operations Research and Computer  Science for his book "Neuro-Dynamic Programming" (co-authored with John Tsitsiklis), the 2000 Greek National Award for Operations  Research, and the 2001 ACC John R. Ragazzini Education Award. In  2001, he was elected to the United States National Academy of  Engineering.

Dr. Bertsekas' recent books are "Dynamic Programming and Optimal Control: 3rd Edition" (2007), "Introduction to Probability: 2nd Edition" (2008), and Convex Optimization Theory (2009), all published by Athena Scientific.

Besides his professional activities, Professor Bertsekas is interested in travel, portrait, and landscape photography. His pictures have been exhibited on several occasions at M.I.T., and can also be accessed from his www site.