Computer experiments: design, modeling and integration

Computer experiment;Gaussian process models;Metamodels;Surrogate models;Design of experiments;Orthogonal arrays;Structural equation model;Latent variable model;Data center

Qian, Zhiguang ; Industrial and Systems Engineering

University:Georgia Institute of Technology

Department:Industrial and Systems Engineering

关键词: Computer experiment;    Gaussian process models;    Metamodels;    Surrogate models;    Design of experiments;    Orthogonal arrays;    Structural equation model;    Latent variable model;    Data center;   

Others  :  https://smartech.gatech.edu/bitstream/1853/11480/1/zhiguang_qian_200608_phd.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

The use of computer modeling is fast increasing in almost everyscientific, engineering and business arena. This dissertationinvestigates some challenging issues in design, modeling andanalysis of computer experiments, which will consist of four majorparts. In the first part, a new approach is developed to combinedata from approximate and detailed simulations to build asurrogate model based on some stochastic models. In the secondpart, we propose some Bayesian hierarchical Gaussian processmodels to integratedata from different types of experiments. Thethird part concerns the development of latent variable models forcomputer experiments with multivariate response with applicationto data center temperature modeling. The last chapter is devotedto the development of nested space-filling designs for multipleexperiments with different levels of accuracy.