Robust Simulation-optimization Methods Using Kriging Metamodels
نویسندگان
چکیده
Acknowledgements Now that I become closer and closer to the end of this adventure, I would like to deeply thank some people who helped me — in different ways and at different stages of my PhD — to achieve this goal. First of all, I want to thank my advisors: Carlo Meloni, for having introduced me to the fascinating world of research, for his being always so supportive to me, and for his willingness to discuss some research issues, going on until a suitable solution was found. Tiziano Politi, for having provided me the important background I needed to tackle the topic that, afterwards, evolved into the present dissertation. Jack Kleij-nen, for having given me the opportunity to visit Tilburg University and personally collaborate with him; thanks to our intensive and productive meetings, I learned so much during the year I spent working there. I would also like to thank my former roommate , Wim Van Beers, with whom I shared the office during my stay in Tilburg: he was always available to help me, no matter whether he was busy or not; I will never forget what he told me one day I felt a bit frustrated for my research having temporarily come to a stop: " Keep smiling! ". I will... Many thanks also to Dick den Hertog and Inneke Van Nieuwenhuyse for the stimulating and thorough discussions which definitely improved the outcome of my work. Besides the results discussed in this thesis, I cooperated in some research projects at the Politecnico di Bari: therefore, I wish to thank Alessandro Rizzo e Paolo Lino, who — together with Carlo Meloni — have been my first research group and initiated me into my PhD, always making me feel at ease. I am grateful to the referees of my dissertation, Alessandro Agnetis, Bill Biles and Renato De Leone, who gave me valuable remarks to improve the quality of my PhD thesis; their comments really encouraged me to keep on working on this topic. Finally, I want to thank my parents, for their discreet support and for giving me all I needed to concentrate on work whenever I had to. Thanks to my friends for their perseverance in inviting me to have fun together, despite all the times I answered 'no'. In the end, I heartily thank my fiancé, Giulio, for his infinite love, his support and encouragement and for always …
منابع مشابه
Simulation-optimization via Kriging and bootstrapping: a survey
This survey considers the optimization of simulated systems. The simulation may be either deterministic or random. The survey reflects the author’s extensive experience with simulationoptimization through Kriging (or Gaussian process) metamodels using a frequentist (non-Bayesian) approach. The analysis of Kriging metamodels may use bootstrapping. The survey discusses both parametric bootstrappi...
متن کاملBlind Kriging: A New Method for Developing Metamodels
Kriging is a useful method for developing metamodels for product design optimization. The most popular kriging method, known as ordinary kriging, uses a constant mean in the model. In this article, a modified kriging method is proposed, which has an unknown mean model. Therefore it is called blind kriging. The unknown mean model is identified from experimental data using a Bayesian variable sel...
متن کاملRobust Nonconvex Optimization for Simulation-based Problems
In engineering design, an optimized solution often turns out to be suboptimal, when implementation errors are encountered. While the theory of robust convex optimization has taken significant strides over the past decade, all approaches fail if the underlying cost function is not explicitly given; it is even worse if the cost function is nonconvex. In this work, we present a robust optimization...
متن کاملComputationally Inexpensive Metamodel Assessment Strategies
Inmany scienti c and engineering domains, it is common to analyze and simulate complex physical systems using mathematicalmodels.Althoughcomputing resources continue to increase inpower and speed, computer simulation and analysis codes continue togrowincomplexityandremain computationally expensive, limiting their use indesign and optimization. Consequently, many researchers have developed diff...
متن کاملSensitivity Analysis of Simulation Models
This contribution presents an overview of sensitivity analysis of simulation models, including the estimation of gradients. It covers classic designs and their corresponding (meta)models; namely, resolution-III designs including fractional-factorial twolevel designs for rst-order polynomial metamodels, resolution-IV and resolution-V designs for metamodels augmented with twofactor interactions,...
متن کاملRobust Optimization for Unconstrained Simulation-Based Problems
In engineering design, an optimized solution often turns out to be suboptimal, when errors are encountered. While the theory of robust convex optimization has taken significant strides over the past decade, all approaches fail if the underlying cost function is not explicitly given; it is even worse if the cost function is nonconvex. In this work, we present a robust optimization method, which ...
متن کامل