Surrogate Models / Reduced Order Modeling
We present various methods for interpolating the output of expensive computer models with cheaper reduced order models. We examine methods statistical models such as Kriging/Gaussian process emulators as well as reduced basis methods.
Documents
Homework3.pdf | Homework 3 - Due May 31. |
hw3.tgz | Codes for Homework 3.. |
pmpack.tgz | A suite of Matlab tools for working orthogonal polynomials and quadrature. |
outline14.pdf | Handout for lecture 14. |
References
- A Taxonomy of Global Optimization Methods Based on Response Surfaces. Donald Jones.
- Computer Experiments. Art Owen and James Koehler.
- Model Reduction for Large-Scale Systems with High-Dimensional Parametric Input Space. Tan Bui-Thanh, Karen Willcox, and Omar Ghattas.