Come join ISE’s faculty and fellow students as they welcome Barry Nelson from Northwestern University as he discusses how Parallel Adaptive Survivor Selection (PASS) jettisons the standard framework of ranking & selection to produce a new class of algorithms.
As always refreshments will be served from 10:30 – 10:50 am in room 428 Daniels Hall.
Stochastic computer simulation is used to design and improve real-world systems in settings as diverse as hospital staffing, manufacturing system design, portfolio planning, supply chain management, and telecommunication network evaluation. The class of methods known collectively as “ranking & selection” have been a theoretical and practical success story for finding the best among a finite number of simulated systems (“feasible solutions”) with a statistical guarantee of correctness. However, the tricks that made these methods efficient in terms of the number of simulated observations can also render them (sometimes horribly) inefficient and even invalid when implemented in a parallel, high-performance computing environment, which is of course exactly what one wants to do. Parallel adaptive survivor selection (PASS) jettisons the standard framework of ranking & selection to produce a new class of algorithms that actually benefit statistically from a large number of simulated systems without any increase in coordination among parallel processors or added conservatism. The basic theory of PASS and the results of large-scale experiments will be presented. No background in ranking & selection or stochastic simulation is assumed.
Barry L. Nelson is the Walter P. Murphy Professor of the Department of Industrial Engineering and Management Sciences at Northwestern University and a Distinguished Visiting Scholar at Lancaster University in England. His research focus is on the design and analysis of computer simulation experiments on models of discrete-event, stochastic systems, including methodology for simulation optimization, quantifying and reducing model risk, variance reduction, output analysis, metamodeling and multivariate input modeling. He has published numerous papers and three books, including Foundations and Methods of Stochastic Simulation: A First Course (Springer, 2013). Nelson is a Fellow of INFORMS and IIE. Further information can be found at www.iems.northwestern.edu/~nelsonb/.
