Abstract: Suppose that we have to make a decision in real time where we have to choose between a number of alternative ways of operating a system. A performance measure is available for comparing the alternatives. We also have a simulation model and can make runs of this to estimate the performance measure, but not all alternatives can be examined because there is a strict limit to the time available for simulations. A balance is needed between making a few long runs in which the performance measure is accurately estimated but only for a few alternatives, and making a large number of short runs but with the performance measure poorly estimated. We analyse how the run length should be selected to ensure that an alternative with a good performance measure will be found with high probability. We give a real example arising in the provision of fire service emergency cover.
Bio: Russell C. H. Cheng is Emeritus Professor of Operational Research at the University of Southampton. He has an M.A. and the Diploma in Mathematical Statistics from Cambridge University, England. He obtained his Ph.D. from Bath University. He is a former Chairman of the U.K. Simulation Society, a Fellow of the Royal Statistical Society and Fellow of the Institute of Mathematics and Its Applications. His research interests include: design and analysis of simulation experiments and parametric estimation methods. He was a Joint Editor of the IMA Journal of Management Mathematics. His email and web addresses are email@example.com and http://www.personal.soton.ac.uk/rchc.