In many Markov modeling contexts, the system under consideration exhibits strong time-of-day effects, day-of-week effects, or seasonality effects. In fact, most real-world systems that are modeled as Markov processes exhibit such non-stationarities. Nevertheless, the great majority of the academic literature focuses on modeling and theory for Markov processes with stationary transition probabilities, which describe systems that have no such time-of-day effects or seasonality. In this talk, we will describe an analytical approximation approach to handle non-stationary Markov models, anchored on traditional stationary analysis, that are available when the transition probabilities are slowly changing or rapidly changing over the time horizon in question. We will then discuss the use of Monte Carlo simulation to study such non-stationary processes, followed by implications on decision-making under non-stationarities. Based on joint work with Peter W. Glynn, Harsha Honnappa.

Bio:  Zeyu Zheng is an assistant professor in the Department of Industrial Engineering & Operations Research at UC Berkeley. He recently received a Ph.D. in management science and engineering and a Ph.D. minor in statistics from Stanford University. Previously, he earned a masters degree in economics from Stanford, and a B.S. in mathematics from Peking University. His research interests include simulation, data analytics, stochastic modeling, and service operations management.