Skip to content Skip to navigation

Academic Events

Academic Seminar on From Continuous to Discrete, and Back -- Optimality Conditions in Dynamic Robust Optimization

On Mar 1, Associate Professor Bai Yang attended an academic Seminar held in Industrial & Enterprise Systems Engineering (ISE) at 303 Transportation Building. Associate Professor Bai is now visiting University of Illinois at Urbana-Champaign (UIUC) as visiting scholar. He is now working on Markov decision process (MDP) as well as energy security policy modelling.

Professor Dan A Iancu gave a speech on Dynamic Robust Optimization. Classical paradigms for decision making under very limited information - such as robust optimization (RO) - typically require optimizing worst-case outcomes. Such robust criteria are often degenerate, admitting multiple optimal solutions. This study first discusses how this fundamental property can be used to derive necessary and sufficient conditions for the optimality of specific classes of policies in dynamic RO models. This generalizes several disparate results in the literature, and draws interesting connections between RO and combinatorial optimization (particularly lattice programming and discrete convexity) and global optimization (the theory of concave envelopes), which may be of independent interest. The author also shows how the conditions can be used to recover optimal policies in several applications of interest, including multi-period supply chain contracting models, healthcare monitoring, and collateralized lending. The optimal policies have a simple and intuitive structure, and can be found by solving scalable optimization problems, even when traditional approaches such as Dynamic Programming fail.

Dan Iancu is an Associate Professor of Operations, Information and Technology at the Stanford Graduate School of Business. A native of Romania, Professor Iancu holds a B.S. degree in Electrical Engineering and Computer Science from Yale University, an S.M. in Engineering Sciences from Harvard University, and a Ph.D. in Operations Research from the Sloan School of Management at MIT. Prior to joining Stanford, he spent one year as a Goldstine Fellow in the Risk Analytics Group at the IBM T.J. Watson Research Center. His research interests focuses on dynamic optimization under uncertainty, with applications in problems at the interface of operations, finance, and risk management. He was the recipient of several best paper awards (INFORMS JFIG 2013, INFORMS Optimization Society 2009), and of teaching prizes at Harvard and MIT Sloan.