Course Descriptions
Design and Analysis of Computer Experiments
This course will cover statistical methods for dealing with
experiments in which a deterministic computer model is the primary object of
study.
Topics to be covered include
- An introduction to computer models
- Sampling-based methods of sensitivity and uncertainty analysis
- Gaussian stochastic process (GaSP) meta-models
- Experimental designs for GaSP models
- Empirical Bayes treatment of GaSP models
- Full Bayes treatment of GaSP models
- Generalizations for multivariate outputs
- Model validation
- Model calibration
- Inverse problems and response optimization
The course will be primarily designed for statistics graduate
students, and graduate students in other fields who use computer models and
have advanced statistics background. Students should have completed the MS core
as prerequisite material.
|
