Course Level: Intermediate
Format Requirements: Participants will get hands-on experience while learning the analytical foundations of these techniques. They will learn how to use Arena’s Process Analyzer and OptQuest for performing simulation optimization, and learn the concept of variance reduction and common random numbers (CRN) through some Excel examples. Participants should bring a laptop computer that can run Windows 2000 or higher. We will use Excel as well as an academic version of Arena, which will be provided for download prior to the course. Not that this version is not compatible with Windows 7 64 bits, and if you have a Mac, you will need to run some type of windows emulator to use Arena.
Background: This course is intended for people with some familiarity in discrete-event or Monte Carlo simulation, and who wish to learn about simulation optimization and variance reduction techniques (participants of the morning SMDM short course on "Introduction to Discrete-Event Simulation" may wish to take this as a follow-up course the same day). Applications include both system-level models (e.g., capacity planning, scheduling) as well as patient-level models (e.g., disease progression and treatment timing).
Description and Objectives: The objectives of this course are to:
- Learn generally about what makes simulation optimization challenging and what methods are available for finding “the best” solution(s)
- Learn how Arena’s Process Analyzer can help identify the best scenario or a subset containing the best with high probability when a moderate number of alternatives exist
- Learn how Arena’s OptQuest tool can help search for good scenarios when thousands of scenarios may be feasible
- Learn how comparing scenarios may become significantly more precise through the use of CRN, and to understand when implementing CRN may or may not be worth it.