Design and Analysis of Simulation Experiments (eBook)

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.  

Preface 7
Contents 10
Introduction 13
What is simulation? 13
What is DASE? 19
DASE symbols and terms 22
Solutions for exercises 24
Low-order polynomial regression metamodels and their designs: basics 26
Introduction 27
Linear regression analysis: basics 30
Linear regression analysis: first-order polynomials 38
First-order polynomial with a single factor 38
First-order polynomial with several factors 39
Designs for first-order polynomials: resolution-III 47
2k-p designs of resolution-III 47
Plackett-Burman designs of resolution-III 50
Regression analysis: factor interactions 51
Designs allowing two-factor interactions: resolution-IV 53
Designs for two-factor interactions: resolution-V 57
Regression analysis: second-order polynomials 60
Designs for second-degree polynomials: Central Composite Designs (CCDs) 61
Optimal designs and other designs 62
Validation of metamodels 65
Coefficients of determination and correlation coefficients 65
Cross-validation 68
More simulation applications 74
Conclusions 77
Appendix: coding of nominal factors 77
Solutions for exercises 80
Classic assumptions revisited 83
Introduction 83
Multivariate simulation output 84
Designs for multivariate simulation output 87
Nonnormal simulation output 88
Realistic normality assumption? 88
Testing the normality assumption 89
Transformations of simulation I/O data, jackknifing, and bootstrapping 90
Heterogeneous simulation output variances 97
Realistic constant variance assumption? 97
Testing for constant variances 98
Variance stabilizing transformations 99
LS estimators in case of heterogeneous variances 99
Designs in case of heterogeneous variances 102
Common random numbers (CRN) 103
Realistic CRN assumption? 104
Alternative analysis methods 104
Designs in case of CRN 106
Nonvalid low-order polynomial metamodel 107
Testing the validity of the metamodel 107
Transformations of independent and dependent regression variables 108
Adding high-order terms to a low-order polynomial metamodel 108
Nonlinear metamodels 109
Conclusions 109
Solutions for exercises 110
Simulation optimization 111
Introduction 111
RSM: classic variant 115
Generalized RSM: multiple outputs and constraints 120
Testing an estimated optimum: KKT conditions 126
Risk analysis 133
Latin Hypercube Sampling (LHS) 136
Robust optimization: Taguchian approach 140
Case study: Ericsson's supply chain 145
Conclusions 147
Solutions for exercises 148
Kriging metamodels 149
Introduction 149
Kriging basics 150
Kriging: new results 157
Designs for Kriging 159
Predictor variance in random simulation 161
Predictor variance in deterministic simulation 162
Related designs 164
Conclusions 165
Solutions for exercises 166
Screening designs 167
Introduction 167
Sequential Bifurcation 170
Outline of simplest SB 170
Mathematical details of simplest SB 175
Case study: Ericsson's supply chain 177
SB with two-factor interactions 179
Conclusions 181
Solutions for exercises 182
Epilogue 183
References 185
Index 221