Stochastic simulation optimization : an optimal computing budget allocation PDF
Preview Stochastic simulation optimization : an optimal computing budget allocation
STOCHASTIC SIMULATION OPTIMIZATION An Optimal Computing Budget Allocation SERIES ON SYSTEM ENGINEERING AND OPERATIONS RESEARCH Editor-in-Chief: Yu-Chi Ho (Harvard University, USA) Associate Editors: Chun-Hung Chen (George Mason University, USA) Loo Hay Lee (National University of Singapore, Singapore) Qianchuan Zhao (Tsinghua University, China) About the Series The series provides a medium for publication of new developments and advances in high level research and education in the fields of systems engineering, industrial engineering, and operations research. It publishes books in various engineering areas in these fields. The focus will be on new development in these emerging areas with utilization of the state-of-the-art technologies in addressing the critical issues. The topics of the series include, but not limited to, simulation optimization, simulation process and analysis, agent-based simulation, evolutionary computation/ soft computing, supply chain management, risk analysis, service sciences, bioinformatics/ biotechnology, auctions/competitive bidding, data mining/machine learning, and robust system design. Vol. 1 Stochastic Simulation Optimization: An Optimal Computing Budget Allocation Authors: C.-H. Chen and L. H. Lee Gregory - Stochastic Simulation Optimization.pm2d 8/2/2010, 6:20 PM System Engineering and Operations Research – Vol. 1 STOCHAST IC SIMULAT ION OPT IMIZAT ION An Optimal Computing Budget Allocation Chun-Hung Chen George Mason Univ., USA National Taiwan Univ. Loo Hay Lee National Univ. of Singapore World Scientific NEW JERSEY • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TAIPEI • CHENNAI Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. System Engineering and Operations Research — Vol. 1 STOCHASTIC SIMULATION OPTIMIZATION An Optimal Computing Budget Allocation Copyright © 2011 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher. For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher. ISBN-13 978-981-4282-64-2 ISBN-10 981-4282-64-2 Typeset by Stallion Press Email: [email protected] Printed in Singapore. Gregory - Stochastic Simulation Optimization.pm1d 8/2/2010, 6:20 PM May11,2010 6:45 SPI-B930 9inx6in b930-fm Foreword to the WSP Series on System Engineering and Operation Research Advancement in science and technology often blurs traditional dis- ciplinary boundary. Control system theory and practice, operations research, and computational intelligence combine to contribute to moderncivilization inmyriadways.From trafficcontrol onland,sea, and air, to manufacturing automation, to social and communication networks, these knowledge-based and human-made systems rely on research results in the above disciplinary topics for their smooth and efficient functioning. TheWorldScientificPublishingSeriesonSystemEngineeringand Operations Research is launched to fill this niche for students and scholars doing research in these areas. The first book in this series is by two leading scientists in the area of efficent simulation and mod- eling of complex systems. They articulate clearly the computational burden involved in such study and device innovative methodology to overcome the difficulties. I welcome their contribution in inaugurat- ing this series and look forward to additional books in this genre. Yu-Chi Ho Editor-in-Chief WSP Series on System Engineering and Operations Research January 20, 2010 v May11,2010 6:45 SPI-B930 9inx6in b930-fm TThhiiss ppaaggee iinntteennttiioonnaallllyy lleefftt bbllaannkk May11,2010 6:45 SPI-B930 9inx6in b930-fm Preface “Simulation” and “optimization” are two very powerful tools in sys- tems engineering and operations research. With the advance of new computing technology, simulation-based optimization is growing in popularity. However, computational efficiency is still a big concern because (i) in the optimization process, many alternative designs must be simulated; (ii) to obtain a sound statistical estimate, a large number of simulation runs (replications) is required for each design alternative. A usermay beforced to compromise on simulation accu- racy, modeling accuracy, and the optimality of the selected design. There have been several approaches developed to address such an efficiency issue. This book is intended to offer a different but ambi- tious approach by trying to answer the question “what is an optimal (or the most efficient) way to conduct all the simulations in order to find a good or optimal solution (design)?” The ultimate goal is to minimize the total simulation budget while achieving a desired optimality level, or to maximize the probability of finding the best design using a fixed computing budget. The primary idea is called Optimal Computing Budget Allocation (OCBA). This book aims at providing academic researchers and industrial practitioners a comprehensive coverage of the OCBA approach for stochastic simulation optimization. Chapter 1 introduces stochas- vii May11,2010 6:45 SPI-B930 9inx6in b930-fm viii SSO: An Optimal Computing Budget Allocation tic simulation optimization and the associated issue of simulation efficiency. Chapter 2 gives an intuitive explanation of computing budget allocation and discusses its impact on optimization perfor- mance. Then a series of OCBA approaches developed for various problems are presented, from selecting the best design (Chapter 3), selecting a set of good enough designs (Chapter 5), to optimization with multiple objectives (Chapter 6). Chapter 4 provides numeri- cal illustrations, showing that the computation time can be reduced significantly. Chapter 4 also offers guidelines for practical implemen- tation of OCBA. Chapter 7 extends OCBA to large-scale simulation optimization problems. The OCBA technique is generic enough that it can be integrated with many optimization search algorithms to enhance simulation optimization efficiency. Several potential search techniques are explored. Finally, Chapter 8 gives a generalized view of the OCBA framework, and shows several examples of how the notion of OCBA can be extended to problems beyond simulation and/or optimization such as data envelopment analysis, experiments of design, and rare-event simulation. To help those readers without much simulation background, in the appendix, we offer a short but comprehensive presentation of stochastic simulation basics. We also include a basic version of OCBA source code in the appendix. OCBA has several strengths: it is effective, easy to understand, simpletoimplement,andcanbeeasilygeneralized orintegratedwith other methods to extend its applicability. We believe that this book is highly useful for different purposes. 1. For researchers, this book offers a series of promising approaches for efficiency enhancement in computer simulation, stochastic optimization, statistical sampling, and ranking and selection. The generalized framework may lead to numerous new lines of researches. 2. For courses, this book could serve as a textbook for advanced stochastic simulation or simulation optimization courses. They can cover Appendix A, Chapters 1 and 2 for introductions; Chap- ters 3 through 4, and parts of Chapters 5 through 8 for advanced materials. May11,2010 6:45 SPI-B930 9inx6in b930-fm Preface ix 3. For practioners, this book offers a simple but effective approach to enhance the computational efficiency of simulation optimiza- tion. Simulation practioners from industries, governments, and the military should find this book useful and relatively easy to read and apply because it gives numerous intuitive illustrations, well-structured algorithms, and practical implementation guide- lines.
The list of books you might like
