Design and Analysis of Simulation Experiments

This is an advanced expository book on statistical methods for the Design and Analysis of Simulation Experiments (DASE). Though the book focuses on DASE for discrete-event simulation (such as queuing and inventory simulations), it also discusses DASE for deterministic simulation (such as engineering and physics simulations). The text presents both classic and modern statistical designs. Classic designs (e.g., fractional factorials) assume only a few factors with a few values per factor. The resulting input/output data of the simulation experiment are analyzed through low-order polynomials, which are linear regression (meta)models. Modern designs allow many more factors, possible with many values per factor. These designs include group screening (e.g., Sequential Bifurcation, SB) and space filling designs (e.g., Latin Hypercube Sampling, LHS). The data resulting from these modern designs may be analyzed through low-order polynomials for group screening and various metamodel types (e.g., Kriging) for LHS.

In this way, the book provides relatively simple solutions for the problem of which scenarios to simulate and how to analyze the resulting data.

The book also includes methods for computationally expensive simulations. It discusses only those tactical issues that are closely related to strategic issues; i.e., the text briefly discusses run-length and variance reduction techniques.

The leading textbooks on discrete-event simulation pay little attention to the strategic issues of simulation. The author has been working on strategic issues for approximately forty years, in various scientific disciples--such as operations research, management science, industrial engineering, mathematical statistics, economics, nuclear engineering, computer science, and information systems.

The intended audience is comprised of researchers, graduate students, and mature practitioners in the simulation area. They are assumed to have a basic knowledge of simulation and mathematical statistics; nevertheless, the book summarizes these basics, for the readers' convenience.

Autorentext

Jack Kleijnen is well known internationally for being a leading researcher in simulation for more than 30 years. He is the author of highly cited books in the area of statistical techniques in simulation that were published between 1974 and 1992. He is an excellent writer and researcher, and hence, ideally suited to write this important book for the field.

On 25 February 2008 Her Majesty Beatrix, Queen of the Netherlands, appointed Jack Kleijnen a Knight in the Order of the Netherlands Lion.

Klappentext

This is an advanced expository book on statistical methods for the Design and Analysis of Simulation Experiments (DASE). Though the book focuses on DASE for discrete-event simulation (such as queuing and inventory simulations), it also discusses DASE for deterministic simulation (such as engineering and physics simulations). The text presents both classic and modern statistical designs. Classic designs (e.g., fractional factorials) assume only a few factors with a few values per factor. The resulting input/output data of the simulation experiment are analyzed through low-order polynomials, which are linear regression (meta)models. Modern designs allow many more factors, possible with many values per factor. These designs include group screening (e.g., Sequential Bifurcation, SB) and space filling designs (e.g., Latin Hypercube Sampling, LHS). The data resulting from these modern designs may be analyzed through low-order polynomials for group screening and various metamodel types (e.g., Kriging) for LHS.

In this way, the book provides relatively simple solutions for the problem of which scenarios to simulate and how to analyze the resulting data.

The book also includes methods for computationally expensive simulations. It discusses only those tactical issues that are closely related to strategic issues; i.e., the text briefly discusses run-length and variance reduction techniques.

The leading textbooks on discrete-event simulation pay little attention to the strategic issues of simulation. The author has been working on strategic issues for approximately forty years, in various scientific disciples--such as operations research, management science, industrial engineering, mathematical statistics, economics, nuclear engineering, computer science, and information systems.

The intended audience is comprised of researchers, graduate students, and mature practitioners in the simulation area. They are assumed to have a basic knowledge of simulation and mathematical statistics; nevertheless, the book summarizes these basics, for the readers' convenience.

Zusammenfassung

Simulation is a widely used methodology in all the Applied Science disciplines. It provides a flexible, powerful and intuitive tool for investigating how to design a process or system, and how to maximize its efficiency. Simulation's effectiveness depends greatly on how well the simulation experiments are designed and analyzed. Design and Analysis of Simulation Experiments will focus on this crucial phase in the overall process of applying simulation. The book will include the best of both classic and modern methods of simulation experimentation, and as a result, it will provide a state-of-the-art treatment of the topic. This book will be the standard reference book on the topic for both researchers and sophisticated practitioners in the area, and it will be used as a textbook in courses or seminars focusing on this topic.

Inhalt

Preface.- Introduction.- Black Box Metamodels.- Low-Order Polynomial Regression.- Metamodels and Designs: A Single Factor.- Low-Order Polynomial Models and Designs: Multiple Factors.- Low-Order Polynomial Models and Screening Designs: Hundreds of Factors.- Kriging Metamodels.- Latin Hypercube Sampling (LHS) and other Space-Filling Designs.- Cross-Validation of Metamodels.- Conclusions and Further Research.