Seminar by Dr. Yongpei Guan

Interim Chair and Professor – Industrial and Systems Engineering

Director – Computational Stochastic Optimization Lab

University of Florida

Wednesday, October 7, 2015 at 4:00 – 5:00 pm

255 Townshend Hall, 1885 Neil Ave. Mall

The traditional two-stage stochastic programming approach assumes the distribution of the random parameter in a problem is known. In most practices, however, the distribution is actually unknown. Instead, only a series of historic data are available. In this paper, we develop a data-driven stochastic optimization approach to providing a risk-averse decision making under uncertainty. In our approach, starting from a given set of historical data, we first construct a confidence set for the unknown probability distribution utilizing a family of zeta-structure probability metrics. Then, we describe the reference distributions and solution approaches to solving the developed two-stage risk-averse stochastic program, corresponding to the given set of historical data, for the cases in which the true probability distributions are discrete and continuous, respectively. More specifically, for the case in which the true probability distribution is discrete, we reformulate the risk-averse problem to a traditional two-stage robust optimization problem. For the case in which the true probability distribution is continuous, we develop a sampling approach to obtaining the upper and lower bounds of the risk-averse problem, and prove that these two bounds converge to the optimal objective value uniformly at the sample size increases. Furthermore, we prove that, for both cases, the risk-averse problem converges to the risk-neutral one as more data samples are observed. Finally, the experiment results on newsvendor and facility location problems show how numerically the optimal objective value of the risk-averse stochastic program converges to the risk-neutral one, which indicates the value of data. This is joint work with Chaoyue Zhao.

Yongpei Guan currently serves as Interim Chair and Professor of Industrial and Systems Engineering and Director of the Computational Stochastic Optimization Lab at the University of Florida. His research interests include nonparametric statistical estimation and stochastic optimization, discrete optimization, and stochastic impulse control with their applications in supply chain management and power system analysis with renewable energy integration. His works in these areas have led to NSF Career Award 2008 and Office of Naval Research Young Investigator Award 2010, and have been published in IEEE Transactions on Power Systems, Mathematical Programming, and Operations Research. His Ph.D. students have won the Nicholson Best Student Paper Award (first place) from INFORMS and Pritsker Doctoral Dissertation Awards (second and third places) from IIE. He is currently the associate editor for Journal of Global Optimization and Computational Optimization and Applications, as well as the newsletter editor for the INFORMS Computing Society. He was also nominated and served as the chair of the 2014 IIE Annual Conference ISERC Program, and invited and served as the 2013 Guest Editor-in-Chief for the Special Issue on “Optimization Methods and Algorithms Applied to Smart Grid” for IEEE Transactions on Smart Grid. Yongpei Guan obtained his Ph.D. from Georgia Tech in 2005.