This course will discuss practical and easy-to-understand statistical techniques for comparing alternative system designs, variance-reduction techniques for obtaining more precise simulation results for the same amount of computing, the use of experimental design techniques to determine important system factors, and simulation-based optimization. The course will also present an introduction to agent-based modeling and simulation, which is arguably the "hottest" topic in simulation modeling today. All concepts will be illustrated by one or more examples or case studies. The prerequisite for this seminar is the "Fundamental Principles" course, or the equivalent knowledge..
Versions of this seminar have been given for organizations such as AT&T, Boeing, Joint Warfare Analysis Center, Nortel Networks, Sandia National Labs, U.S. Air Force, U.S. Army, U.S. Forces Korea, U.S. Navy, and Xerox.
Each attendee who is taking this seminar as a stand-alone course will receive a copy of the book Simulation Modeling and Analysis (4th Edition) by Averill M. Law as part of their registration fee – this book is widely considered to be the “bible” of the simulation industry.
What You Will Learn:
1. Advanced Techniques for Output-Data Analysis
Determining the number of simulation runs (replications) required to estimate a mean with a specified precision (e.g., with no more than 5 percent error)
Estimating probabilities and percentiles
Dealing with multiple measures of performance
2. Comparing Alternative System Designs
Confidence intervals for comparing two or more systems
Procedures to select the best system with a specified probability
3. Variance-Reduction Techniques
Using the method of common random numbers to obtain more precise
“Tricks” for synchronizing random numbers across different system designs
Multiple random-number streams
Generation of all entity attributes upon arrival
“Wasting” random numbers
Use of the inverse-transform technique for generating random values
4.
Design of Experiments for Simulation ModelingHow experimental design can identify key system factors
Factorial
Response-surface designs for prediction and optimization
Failure of classical statistical assumptions and how to address this in simulation modeling
5. Simulation-Based Optimization
How the integration of optimization modules
Available optimization modules and what features to look for
Live optimization of a model with 7 decision variables
6. Introduction to Agent-Based Modeling and Simulation
What is agent-based simulation and how can it benefit your organization
Available simulation software and toolkits
Successful applications