WSC 2007 Final Abstracts

Introduction to Simulation
David Goldsman (Georgia Institute of Technology)

Abstract:
This is an introductory tutorial on the statistical aspects of computer simulation, and is intended to serve as a springboard to many of the other introductory tutorials that appear elsewhere in the Proceedings. We present a number of motivational examples, followed by material on random number and random variate generation, input analysis of the random variables that drive a simulation, and output analysis of the random observations that a simulation produces.

Representing and Generating Uncertainty Effectively
W. David Kelton (University of Cincinnati)

Abstract:
Stochastic simulations involve at least some random inputs. This introductory tutorial is meant to call attention to the need to model and generate such inputs in ways that may not be the standard or defaults in simulation-modeling software. There are both dangers involved with doing things inappropriately, as well as opportunities to do things better, making for more accurate and more precise results from simulations. Specific issues include possible dependence across and within random inputs, use of empirical distributions, and non-default use of the underlying random-number generator. Suggestions for novel ways of implementing some of these ideas in simulation-modeling software are offered.

The Optimizing-simulator: Merging Simulation and Optimization Using Approximate Dynamic Programming
Warren B. Powell (Princeton University)

Abstract:
There is a wide range of simulation problems that involve making decisions during the simulation, where we would like to make the best decisions possible, taking into account not only what we know when we make the decision, but also the impact of the decision on the future. Such problems can be formulated as dynamic programs, stochastic programs and optimal control problems, but these techniques rarely produce computationally tractable algorithms. We demonstrate how the framework of approximate dynamic programming can produce near-optimal (in some cases) or at least high quality solutions using techniques that are very familiar to the simulation community. The price of this challenge is that the simulation has to be run iteratively, using statistical learning techniques to produce the desired intelligence. The benefit is a reduced dependence on more traditional rule-based logic.

Fundamentals of Simulation Modeling
Paul Sanchez (Naval Postgraduate School)

Abstract:
We start with basic terminology and concepts of modeling, and decompose the art of modeling as a process. This overview of the process helps clarify when we should or should not use simulation models. We discuss some common missteps made by many inexperienced modelers, and propose a concrete approach for avoiding those mistakes. After a quick review random number and random variate generation, we view the simulation model as a black-box which transforms inputs to outputs. This helps frame the need for designed experiments to help us gain better understanding of the system being modeled.

Introduction to Modeling and Generating Probabilistic Input Processes for Simulation
Michael E. Kuhl (Rochester Institute of Technology), Emily K. Lada (SAS Institute Inc.), Natalie M. Steiger (University of Maine), Mary Ann Wagner (SAIC) and James R. Wilson (North Carolina State University)

Abstract:
Techniques are presented for modeling and generating the univariate probabilistic input processes that drive many simulation experiments. Emphasis is on the generalized beta distribution family, the Johnson translation system of distributions, and the Bezier distribution family. Also discussed are nonparametric techniques for modeling and simulating time-dependent arrival streams using nonhomogeneous Poisson processes. Public-domain software implementations and current applications are presented for each input-modeling technique. Many of the references include live hyperlinks providing online access to the referenced material.

Statistical Analysis of Simulation Output Data: The Practical State of the Art
Averill M. Law (Averill M. Law & Associates)

Abstract:
One of the most important but neglected aspects of a simulation study is the proper design and analysis of simulation experiments. In this tutorial we give a state-of-the-art presentation of what the practitioner really needs to know to be successful. We will discuss how to choose the simulation run length, the warmup-period duration (if any), and the required number of model replications (each using different random numbers). The talk concludes with a discussion of three critical pitfalls in simulation output-data analysis.

Work Smarter, Not Harder: Guidelines for Designing Simulation Experiments
Susan M. Sanchez (Naval Postgraduate School)

Abstract:
We present the basic concepts of experimental design, the types of goals it can address, and why it is such an important and useful tool for simulation. A well-designed experiment allows the analyst to examine many more factors than would otherwise be possible, while providing insights that cannot be gleaned from trial-and-error approaches or by sampling factors one at a time. We focus on experiments that can cut down the sampling requirements of some classic designs by orders of magnitude, yet make it possible and practical to develop a better understanding of a complex simulation model. Designs we have found particularly useful for simulation experiments are illustrated using simple simulation models, and we provide links to other resources for those wishing to learn more. Ideally, this tutorial will leave you excited about experimental designs---and prepared to use them---in your upcoming simulation studies.

Agent-based Modeling and Simulation: Desktop ABMS
Charles M. Macal and Michael J. North (Argonne National Laboratory)

Abstract:
Agent-based modeling and simulation (ABMS) is a new approach to modeling systems comprised of autonomous, interacting agents. ABMS promises to have far-reaching effects on the way that businesses use computers to support decision-making and researchers use electronic laboratories to support their research. Some have gone so far as to contend that ABMS is a third way of doing science, in addition to traditional deductive and inductive reasoning. Computational advances have made possible a growing number of agent-based models across a variety of application domains. Applications range from modeling agent behavior in the stock market, supply chains, and consumer markets, to predicting the spread of epidemics, the threat of bio-warfare, and factors responsible for the fall of ancient civilizations. This tutorial describes theoretical and practical foundations of ABMS, identifies toolkits and methods for developing agent models, and illustrates the development of a simple agent model of shopper behavior using spreadsheets.

Tips for Successful Practice of Simulation
Deborah Sadowski (Rockwell Automation)

Abstract:
Succeeding with a technology as powerful as simulation involves much more than the technical aspects you may have been trained in. The parts of a simulation study that are outside the realm of modeling and analysis can make or break the project. This paper explores the most common pitfalls in performing simulation studies and identifies approaches for avoiding these problems.