Comparisons Among Methods In Design And Modeling For Compute Experiments And Physical Experiments With Model Uncertainty |
Posted on:2011-09-28 | Degree:Master | Type:Thesis |
Country:China | Candidate:W J Wang | Full Text:PDF |
GTID:2190360308984035 | Subject:Basic mathematics |
Abstract/Summary: | PDF Full Text Request |
There are many methods of design and modeling for computer experiments. For comparing different designs and modeling techniques we choose four true mod-els in this paper that have been used in the literature. Three types of design with four different sizes and four modeling techniques are employed in our comparisons. Based on the comparisons, some empirical conclusions are given and we believe these conclusions might provide usefulness in other projects in computer experiments and physical experiments with model unknown. |
Keywords/Search Tags: | artificial neural network, design of experiment, kriging, Latin hyper-cube design, nearest, orthogonal design, second-order polynomial response surface, uniform design, error |
PDF Full Text Request |
Related items |