内容简介：　　　

　　Performance evaluation of increasingly complex human-made systems requires the use of simulation models. However, these systems are difficult to describe and capture by succinct mathematical models. The purpose of this book is to address the difficulties of the optimization of complex systems via simulation models or other computation-intensive models involving possible stochastic effects and discrete choices. This book establishes distinct advantages of the "softer" ordinal approach for search-based type problems, analyzes its general properties, and shows the many orders of magnitude improvement in computational efficiency that is possible.

英文目录：
Preface
Acknowledgements
I Introduction
II Ordinal Optimization Fundamentals
　　1 Two basic ideas of Ordinal Optimization (OO)
　　2 Definitions, terminologies, and concepts for OO
　　3 A simple demonstration of OO
　　4 The exponential convergence of order and goal softening
　　　　4.1 Large deviation theory
　　　　4.2 Exponential convergence w.r.t. order
　　　　4.3 Proof of goal softening
　　　　　　4.3.1 Blind pick
　　　　　　4.3.2 Horse race
　　5 Universal alignment probabilities
　　　　5.1 Blind pick selection rule
　　　　5.2 Horse race selection rule
　　6 Deterministic complex optimization problem and Kolmogorov equivalence
　　7 Example applications
　　　　7.1 Stochastic simulation models
　　　　7.2 Deterministic complex models
　　8 Preview of remaining chapters
III Comparison of Selection Rules
　　1 Classification of selection rules
　　2 Quantify the efficiency of selection rules
　　　　2.1 Parameter settings in experiments for regression functions
　　　　2.2 Comparison of selection rules
　　3 Examples of search reduction
　　　　3.1 Example: Picking with an approximate model
　　　　3.2 Example: A buffer resource allocation problem
　　4 Some properties of good selection rules
　　5 Conclusion
IV Vector Ordinal Optimization
　　1 Definitions, terminologies, and concepts for VOO
　　2 Universal alignment probability
　　3 Exponential convergence w.r.t. order
　　4 Examples of search reduction
　　　　4.1 Example: When the observation noise contains normal distribution
　　　　4.2 Example: The buffer allocation problem
V Constrained Ordinal Optimization
　　1 Determination of selected set in COO
　　　　1.1 Blind pick with an imperfect feasibility model
　　　　1.2 Impact of the quality of the feasibility model on BPFM
　　2 Example: Optimization with an imperfect feasibility model
　　3 Conclusion
VI Memory Limited Strategy Optimization
　　1 Motivation (the need to find good enough and simple strategies)
　　2 Good enough simple strategy search based on OO
　　　　2.1 Building crude model
　　　　2.2 Random sampling in the design space of simple strategies
　　3 Conclusion
VII Additional Extensions of the OO Methodology
　　1 Extremely large design space
　　2 Parallel implementation of OO
　　　　2.1 The concept of the standard clock
　　　　2.2Extensiontonon-Markov cases using second order approximations
　　　　　　2.2.1 Second order approximation
　　　　　　2.2.2 Numerical testing
　　3 Effect of correlated observation noises
　　4 Optimal Computing Budget Allocation and Nested Partition
　　　　4.1 OCBA
　　　　4.2 NP
　　5 Performance order vs. performance value
　　6 Combination with other optimization algorithms
　　　　6.1 Using other algorithms as selection rules in OO
　　　　　　6.1.1 GA+OO
　　　　　　6.1.2 SA+OO
　　　　6.2 Simulation-based parameter optimization for algorithms
　　　　6.3 Conclusion
VIII Real World Application Examples
　　1 Scheduling problem for apparel manufacturing
　　　　1.1 Motivation
　　　　1.2 Problem formulation
　　　　　　1.2.1 Demand models
　　　　　　1.2.2 Production facilities
　　　　　　1.2.3 Inventory dynamic
　　　　　　1.2.4 Summary
　　　　1.3 Application of ordinal optimization
　　　　　　1.3.1 Random sampling of designs
　　　　　　1.3.2 Crude model
　　　　1.4 Experimental results
　　　　　　1.4.1 Experiment 1: 100 SKUs
　　　　　　1.4.2 Experiment 2: 100 SKUs with consideration on satisfaction rate
　　　　1.5 Conclusion
　　2 The turbine blade manufacturing process optimization problem
　　　　2.1 Problem formulation
　　　　2.2 Application of OO
　　　　2.3 Conclusion
　　3 Performance optimization for a remanufacturing system
　　　　3.1 Problem formulation of constrained optimization
　　　　3.2 Application of COO
　　　　　　3.2.1 Feasibility model for the constraint
　　　　　　3.2.2 Crude model for the performance
　　　　　　3.2.3 Numerical results
　　　　3.3 Application of VOO
　　　　3.4 Conclusion
　　4 Witsenhausen problem
　　　　4.1 Application of OO to find a good enough control law
　　　　　　4.1.1 Crude model
　　　　　　4.1.2 Selection of promising subsets
　　　　4.2 Application of OO for simple and good enough control laws
　　　　4.3 Conclusion
Appendix A Fundamentals of Simulation
　　1 Introduction to simulation
　　2 Random numbers and variables generation
　　　　2.1 The linear congruential method
　　　　2.2 The method of inverse transform
　　　　2.3 The method of rejection
　　3 Sampling, the central limit theorem, and confidence intervals
　　4 Nonparametric analysis and order statistics
　　5 Additional problems of simulating DEDS
　　6 The alias method of choosing event types
Appendix B Introduction to Stochastic Processes and Generalized Semi-Markov Processes as Models for Discrete Event Dynamic Systems and Simulations
　　1 Elements of stochastic sequences and processes
　　2 Modeling of discrete event simulation using stochastic sequences
Appendix C Universal Alignment Tables for the Selection Rules in Chapter III
Appendix D Exercises
　　1 True/False questions
　　2 Multiple-choice questions
　　3 General questions
References
Index