An illustration of the Lorenz attractor

I am a Electrical and Computer Engineering doctoral student at the University of Illinois. My research is directed towards finding and analyzing optimal (or nearly optimal) policies for partially observed Markov decision processes (POMDPs).

My interests lie in the intersection of random processes, control theory, robotics, and AI. In particular, I am interested in the impact "information" has on on both executing and learning policies for partially observed systems. I wish to address these questions for stochastic systems via analyzing the sensitivity of policies due to variation in the available information. This analysis should endow us with an understanding of the role of information as well as form the basis of computational methods to adaptively learn optimal policies.