Please use this identifier to cite or link to this item: 192.168.6.56/handle/123456789/44182
Title: Stochastic Simulation Optimization
Authors: Chun-Hung Chen Loo Hay Lee
Keywords: Systems engineering
Issue Date: 2011
Publisher: World Scientific
Description: 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 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 performance. 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 numerical illustrations, showing that the computation time can be reduced significantly. Chapter 4 also offers guidelines for practical implementation 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
URI: http://10.6.20.12:80/handle/123456789/44182
ISBN: 978-981-4282-64-2
Appears in Collections:Regional and Local Development Studies

Files in This Item:
File Description SizeFormat 
26.pdf2.12 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.