Stochastic Kriging for Simulation on Demand

Date:  March 25, 2009
Time:  11:00 a.m.
Place:  3210 Shelby Center

Barry Nelson
Charles Deering McCormick Professor of Teaching Excellence and Chair
Department of Industrial Engineeirng and Management Sciences
Northwestern University

Biography

Nelson's research focus includes the design and analysis of computer simulation experiments on models of discrete-event, stochastic systems and applications in manufacturing, services, finance and transportation. He is a fellow of the Institute for Operations Research and the Management Sciences (INFORMS). In 2006, he received the Outstanding Simulation Publication Award from the INFORMS Simulation Society for his work on simulation optimization. Nelson has published numerous papers and two books, including Discrete-Event System Simulation, Fourth edition (Prentice Hall, 2005), which has been adopted by more than 60 universities.

Abstract

Stochastic Kriging for Simulation on Demand
In automobile assembly, semiconductor manufacturing and financial risk management, situations exist where there is value in running extensive stochastic simulation experiments when time is plentiful, so that simulation-quality results can be delivered "on demand" when time is scarce. In his talk, Nelson will describe his work on stochastic kriging, an extension of kriging used in the design and analysis of deterministic computer experiments. Kriging is a group of geostatistical techniques to introduce the value of a random field at an observed location based on observations of its value at nearby locations.