Advances in Modeling and Simulation (eBook)

Dr. Andreas Tolk is a Technology Integrator at The MITRE Corporation, Hampton, VA, USA, and adjunct Professor at Old Dominion University, Norfolk, VA, USA.Dr. John Fowler is the Motorola Professor of Supply Chain Management in the W.P. Carey School of Business at Arizona State University, AZ, Tempe, USA.Dr. Guodong Shao is a Computer Scientist in the Systems Integration Division (SID) of the Engineering Laboratory (EL) at the National Institute of Standards and Technology (NIST), Gaithersburg, MD, USA.Dr. Enver Yücesan is a Professor of Operations Management at INSEAD, Fontainebleau, France.

Foreword 7
Preface 9
Contents 11
1 A Brief Introduction to the Winter Simulation Conference 13
Abstract 13
1.1 Introduction 13
1.2 The History of the Winter Simulation Conference 14
1.3 Structure of Current WSC 22
1.4 Contributions of the Chapters 26
References 28
2 Model Is a Verb 29
Abstract 29
2.1 Introduction 29
2.2 The Steps in a Simulation Study 30
2.3 Data 34
2.4 Modeling 35
2.5 Experimenting 36
Acknowledgements 37
References 37
3 Model Calibration 38
3.1 Introduction 38
3.2 Overview 41
3.3 Direct Calibration 43
3.3.1 Direct Calibration Using Stochastic Approximation 44
3.3.2 Direct Calibration of a Drug Resistance Simulation Model in a Parasitology Study Using Discrete Optimization via Simulation 45
3.4 Bayesian Calibration 48
3.4.1 A Gaussian Process-Based Bayesian Calibration Framework 48
3.4.2 Simultaneous Bayesian Calibration of Physical and Tunable Parameters 52
3.5 Summary 53
References 55
4 Validating Emergent Behavior in Complex Systems 58
4.1 Introduction 58
4.2 Identifying Emergent Behavior 60
4.3 Validating Emergent Behavior 63
4.3.1 Identifying Emergent Behavior 64
4.3.2 Behavior Comparison 64
4.3.3 Initial Experiments 68
4.4 Conclusion 71
References 71
5 Input Model Risk 74
5.1 Introduction 74
5.2 Background and Notation 75
5.3 Input Modeling 76
5.4 Quantification of Input Model Risk 78
5.4.1 A Measure of Input Model Risk 78
5.4.2 Variability of "019AF 80
5.4.3 Propagating Input Model Uncertainty 81
5.4.4 Confidence and Credible Intervals 82
5.4.5 Measures of Contribution and Sample-Size Sensitivity 83
5.5 Hedging Against Input Model Risk 85
5.6 Simulation Optimization Under Input Model Risk 85
5.7 Applications of Input Model Risk Quantification 87
5.8 Conclusions and Future Directions 87
References 89
6 The Evolution of Simulation Languages 92
Abstract 92
6.1 Introduction 92
6.2 Discrete Modeling World Views 93
6.2.1 Event Orientation 93
6.2.2 Process Orientation 95
6.2.3 Object Orientation 97
6.3 System Dynamics 99
6.4 Agent Modeling 100
6.5 Multi-paradigm Modeling 102
6.6 Animation 102
6.7 Model Verification/Validation 103
6.8 Experimentation and Analysis 104
6.9 Simulation Execution 105
6.10 Simulation Applications 105
6.11 Summary 106
References 107
7 A Brief History of Time Warp 108
Abstract 108
7.1 Introduction 108
7.2 The Early History of Time Warp: David Jefferson’s Perspective 109
7.3 Origin of Time Warp at RAND 109
7.4 Collaboration with Henry Sowizral 111
7.5 Conservative Versus Optimistic Synchronization 115
7.6 The Virtual Time Paper 117
7.7 Research with My Students at the University of Southern California 118
7.8 The Time Warp Operating System at Jet Propulsion Laboratory 121
7.9 Jade Simulations 126
7.10 Subsequent Years 126
7.11 My Adventures in Time Warp: Perspectives by Richard Fujimoto 127
7.12 Beginnings 127
7.13 Conservative Versus Optimistic Performance 128
7.14 The Rollback Chip, Virtual Time Machine, and Reverse Execution 130
7.15 Simulation Ensembles 132
7.16 Georgia Tech Time Warp 133
7.17 Analytic Models, Memory, and Load Management 134
7.18 The High-Level Architecture 136
7.19 Pivoting to Federated Simulations 140
7.20 In Conclusion: The Future 141
References 142
8 Design and Analysis of Simulation Experiments: Tutorial 146
8.1 Introduction 146
8.2 Basic Linear Regression and Designs 149
8.2.1 R-III Designs for First-Order Polynomials 151
8.2.2 R-IV Designs for First-Order Polynomials 152
8.2.3 R-V Designs for Two-Factor Interactions 152
8.2.4 CCDs for Second-Degree Polynomials 152
8.3 Classic Assumptions Versus Simulation Practice 153
8.3.1 Multivariate Simulation Output 153
8.3.2 Nonnormal Simulation Output 153
8.3.3 Heterogeneous Variances of Simulation Outputs 154
8.3.4 Common Random Numbers 155
8.3.5 Validation of Metamodels 156
8.4 Factor Screening: Sequential Bifurcation 157
8.4.1 Deterministic Simulation and First-Order Polynomials 158
8.4.2 Random Simulation and Second-Order Polynomials 159
8.5 Kriging Metamodels and Their Designs 159
8.5.1 Ordinary Kriging in Deterministic Simulation 160
8.5.2 Designs for Deterministic Simulation 161
8.5.3 Kriging in Random Simulation 162
8.5.4 Monotonic Kriging 162
8.5.5 Global Sensitivity Analysis 163
8.5.6 Risk Analysis or Uncertainty Analysis 163
8.6 Simulation Optimization 164
8.6.1 Response Surface Methodology 164
8.6.2 Kriging for Optimization 166
8.6.3 Robust Optimization 167
References 168
9 Better Big Data via Data Farming Experiments 170
9.1 Background and Motivation 170
9.2 A `Think Big' Mindset 171
9.2.1 Big Questions 172
9.2.2 Big Data as Model Inputs 173
9.2.3 Big Data as Model Outputs 175
9.2.4 Better Big Data via Large-Scale Experiments 176
9.2.5 Bigger can be Easier: The Nuts and Bolts of Data Farming 179
9.2.6 From Better Big Data to Better Insights 181
9.3 Ramifications for Decision Making 183
9.3.1 Ethics in Simulation 183
9.3.2 Leveling the Playing Field 185
9.4 Looking Forward 185
References 186
10 Bayesian Belief Models in Simulation-Based Decision-Making 191
10.1 Introduction 191
10.2 Bayesian Learning 195
10.2.1 Learning the Mean of a Normal Distribution 195
10.2.2 Learning with Correlated Beliefs 197
10.2.3 Learning in Linear Regression 198
10.3 Decision Models 200
10.3.1 Ranking and Selection 201
10.3.2 Learning in a Linear Model 203
10.3.3 Modeling Extensions 205
10.4 Bayesian Decision Procedures 206
10.4.1 Thompson Sampling 206
10.4.2 Value of Information 208
10.4.3 Theoretical Analysis 214
10.5 Approximate Bayesian Inference 217
10.5.1 Moment-Matching for Binary Censored Observations 218
10.5.2 Applications and Extensions 221
10.6 Closing Remarks 222
References 223
11 Simulation Optimization Under Input Model Uncertainty 228
11.1 Introduction 228
11.1.1 Illustrative Example: Risk of Input Uncertainty 229
11.2 Problem Setting and Fundamental Questions 230
11.3 Relevant Literature 231
11.4 Quantifying Uncertainty of Empirical Optimization 232
11.4.1 Asymptotic Distribution of the Optimality Gap 233
11.4.2 Asymptotic Distribution of the Performance of Solutions 236
11.4.3 Multiple Sources of Stochastic Uncertainty 239
11.4.4 Constructing Confidence Intervals 241
11.4.5 Numerical Illustration: Comparison of Confidence Intervals 243
11.5 Simulation Optimization Under Input Uncertainty 244
11.5.1 Bayesian Risk Optimization 245
11.6 Empirical Optimization Versus Bayesian Risk Optimization 250
11.7 Future Directions 253
References 254
12 Parallel Ranking and Selection 257
12.1 Introduction 257
12.1.1 Problem Setting and Notational Conventions 260
12.1.2 Scope 260
12.2 A Stylized Computational Model for R& S
12.3 Mathematical Formulations of Existing R& S Procedures
12.3.1 Mathematical Formulation of Fixed-Precision Guarantees 264
12.3.2 Mathematical Formulation of Fixed-Budget Guarantees 265
12.3.3 Guarantees Require Standard Assumptions 266
12.4 Computational Formulations of Existing Serial R& S Procedures
12.4.1 Computational Formulation of Fixed-Precision Procedures 269
12.4.2 Computational Formulation of Fixed-Budget Procedures 270
12.5 Parallelization: Efficiency and Validity 271
12.6 Existing Parallel Ranking and Selection Procedures 272
12.6.1 Parallel Fixed-Precision Procedures 272
12.6.2 Parallel Fixed-Budget Procedures 277
12.6.3 Available Implementations of Parallel R& S Procedures
12.7 A Future Research Agenda 278
12.8 WSC 2017 280
References 280
13 A History of Military Computer Simulation 284
13.1 Introduction 284
13.2 War Games Preparing the Way for Simulation 285
13.3 Military Computer Simulation 287
13.3.1 The Computer Mainstreams the War Game 288
13.3.2 Rise of the Analytical Simulations 289
13.3.3 Military Simulation Goes Distributed 291
13.4 The Military Track of the Winter Simulation Conference 297
13.5 Examples for Current Challenges 300
13.5.1 Live, Virtual and Constructive Simulation 300
13.5.2 Web-Based and Cloud-Based Simulation 301
13.5.3 Computational Social Sciences 301
13.5.4 Unmanned Assets 302
13.5.5 Other Emergent Challenges 302
13.6 Concluding Remarks 303
References 304
14 Modeling and Analysis of Semiconductor Manufacturing 307
Abstract 307
14.1 Introduction 307
14.2 The 1980s 308
14.3 The 1990’s 309
14.4 The 2000s 312
14.5 The 2010s and Beyond 316
References 318
15 Social and Behavioral Simulation 320
Abstract 320
15.1 Social and Behavioral Simulation: An Introduction 320
15.1.1 Understanding Human and Social Behavior: Empirical, Analytical, Modeling & Simulations
15.1.2 Caveats in Modeling Human and Social Behavior 322
15.1.3 SBS: Features and Relevance 323
15.1.4 A Note on Emergence 324
15.2 Social and Behavioral Simulation: Methods and Applications 325
15.2.1 Evolution of Methods and Techniques 325
15.2.1.1 The Classics: 326
15.2.1.2 Methods and Techniques at WSC: 327
Cellular Automata, Stochastic Simulation and AI 327
System Dynamics 327
Discrete Event Simulation 327
Agent-Based Modeling and Simulations 328
Hybrid Simulations 328
15.2.2 Major Application Areas Across WSC Papers 328
15.2.2.1 Healthcare and Health Services 329
15.2.2.2 Disease Modeling 329
15.2.2.3 Public Policy 330
15.2.2.4 Social Influence, Opinions, Rumors, and Social Networks 330
15.3 Next for Social and Behavioral Simulation: Frontiers and Opportunities 331
15.3.1 Challenges in SBS: Validation and Verification 331
15.3.2 Methodological and Research Frontiers 332
15.3.2.1 Behavioral Modeling 332
15.3.2.2 Simulation Analytics 333
15.3.2.3 Hybrid Modeling 333
15.3.2.4 Large-Scale Agent-Based Modeling and Simulation 333
15.3.3 Conclusion: Looking Forward for SBS at the WSC 334
References 334
16 Analysis of M& S Literature Published in the Proceeding of the Winter Simulation Conference from 1981 to 2016
Abstract 338
16.1 Introduction 338
16.2 Methodology 340
16.2.1 Validation of the Data Set 341
16.2.2 Defining the Data Set for Analysis 343
16.3 Findings 344
16.3.1 Analysis Based on Authorship and Editors of Proceedings 345
16.3.2 Analysis Based on Authors’ Institutional Affiliations 346
16.3.3 Analysis Based on Authors’ Geographic Location 348
16.3.4 Citation Analysis 349
16.3.5 Analysis Based on Research Area 352
16.3.6 Analysis Based on Funding Body 354
16.4 Conclusion 355
References 356
Index 357