Upcoming Events

Simulation Experiment Design and Analysis: Session II

November 5, 2013 10:30AM to 11:30AM
Presenter 
Jeremy Staum, Northwestern University
Location 
Building 203, Room D120
Type 
Seminar
Series 

Getting More from Your Simulation Model, Faster

Abstract:
Have you been troubled by any of the following questions?

  • Is there much uncertainty about the conclusions of my simulation study due to uncertainty about the model’s input parameters?
  • How can I calibrate my model’s input parameters so the model yields realistic behavior? • How can I use the model to find an optimal decision?
  • Which of my model’s parameters are important, in the sense that reasonable changes to the parameters yield large changes in the model’s output?
  • I’ve already done many simulation runs, with different values for the parameters.
  • How can I approximate the model’s output for new values of the parameters without running a new simulation experiment? (I don’t want to wait for the experiment to run or spend more processor-hours.)

This series of two seminars provides an overview of methods to answer these questions and enhance your simulation-based research. Computational efficiency will be emphasized; the methods are applicable even to computationally intensive simulations with many parameters.

Biography:
Jeremy Staum is Associate Professor and Director of Graduate Studies in the Dept. of Industrial Engineering & Management Sciences, Northwestern University. His research focuses on advanced methods of simulation modeling. He is on mini-sabbatical at Argonne in the Decision and Information Sciences Division for the fall quarter, as part of the Northwestern-Argonne Institute Mini-Sabbatical Program.