Join us in welcoming Hamed Rahimian, an assistant professor from Clemson University, as he discusses stochastic optimization. Alums and friends of the program are always welcome.
Prediction and Prescription: On the Interplay Between Machine Learning and Decision-Making Under Uncertainty
The widespread use of data science in industry, dubbed the data science revolution, clearly has changed the way companies, nonprofits and governments do business. Noticeably, the most significant progress has been made around predictive analytics, e.g., machine learning. Despite advances around predictive analytics across myriad technical fields, prescriptive analytics, e.g., decision-making under uncertainty, has received much less but growing attention. In this talk, Rahimian focuses on stochastic optimization as a tool for decision-making under uncertainty, where classically, uncertain parameters are described solely by independent random parameters, ignoring their dependence on multidimensional side information, which is the premise of predictive analytics. In the first part of the talk, I focus on a popular stochastic programming framework, namely expected-value-constrained programming models, which include chance-constrained programs. In the second part of the talk, I focus on ranking and selection, a popular framework for simulation optimization. In both parts, I describe novel approaches to incorporate side information into the decision-making framework with theoretical guarantees. I illustrate findings on synthetic as well as real data in portfolio optimization and assortment optimization.
Hamed Rahimian is an Assistant Professor in the Department of Industrial Engineering at Clemson University. Before joining Clemson IE, he was a Postdoctoral Research Fellow at Northwestern University. He obtained his Ph.D. in Operations Research from The Ohio State University. He is broadly interested in data-driven decision-making under uncertainty to develop new theories, tools, and algorithms with applications in scarce resource allocation in complex, uncertain environments and under incomplete information. His papers have appeared in several top journals, such as Mathematical Programming, SIAM Journal on Optimization, and Operations Research, and his research has been supported by the U.S. Airforce Office of Scientific Research (AFOSR) and the SC Department of Transportation. He also serves as an Associate Editor for the INFORMS Journal on Computing.
