IEEE AP/MTT/EMC/ED TURKEY CHAPTER SEMINAR SERIES -- SEMINAR 57
08 November 2019 (12:40): IEEE AP/MTT/EMC/ED Turkey Seminar Series (S.57)
Speaker: Asst. Prof. Naci Saldı, Özyeğin University
Topic: "Non-signaling approximations of decentralized stochastic control problems"
Location: Middle East Technical University, Ankara, Turkey
Abstract: Decentralized stochastic control theory studies decisions of agents that are acting collectively based on their local information to optimize a common cost function under stochastic uncertainty. It will be a prominent avenue of research for many years to come as modern control systems are increasingly decentralized and interconnected. Currently available techniques (such as classical dynamic programming, policy iteration, value iteration, linear programming, etc.) in centralized stochastic control does not apply under decentralized and asymmetric information structures. Even for a very simple decentralized stochastic control problem (i.e., Witsenhausen’s counterexample), application of classical methods for obtaining good strategies leads to poor performance. Moreover, we note that computing optimal strategies for decentralized stochastic control problems is in general known to be NP hard. Hence, we need new viewpoints and new approaches that overcome decentralized nature of the problem. In this talk, I will consider non-signaling approximation of finite decentralized stochastic control problems. I will first introduce a hierarchy of control policies that can be classified in an increasing order as randomized policies, quantum-correlated policies, and non-signaling policies. Then, I will establish an approximation of optimal policies for decentralized stochastic control systems via extendible nonsignaling policies. I will show that the distance between extendible non-signaling policies and decentralized policies is small if the extension is sufficiently large. Using this result, I will establish a linear programming (LP) approximation of decentralized stochastic control problems. Finally, I will state an open problem regarding computation of optimal value of quantum-correlated policies.
Bio: Naci Saldi received the B.Sc. and M.S. degrees in Electrical and Electronics Engineering from Bilkent University in 2008 and 2010, respectively and the Ph.D. degree in Department of Mathematics and Statistics from Queen’s University in 2015. He was a postdoctoral researcher at the University of Illinois at Urbana-Champaign before joining the Department of Natural and Mathematical Sciences at Özyeğin University as an Assistant Professor. He is a co-author of the book Finite Approximations in Discrete-Time Stochastic Control, published by Springer. His research interests include stochastic and decentralized control, source coding, mean-field games, and applied probability.
Date and Time
- Date: 08 Nov 2019
- Time: 12:40 PM to 02:30 PM
- All times are Turkey
- Add Event to Calendar
Asst. Prof. Naci Saldı