750—Response Surface Methodology. (3) (Prereq: STAT 701 or 705 or consent of department) Methods for fitting (regression) response surfaces and interpreting them subject to random error. Includes designs and industrial process optimization methods.
Usually Offered: Occasionally in Summers
Purpose: To provide graduate students in Statistics and other disciplines with methodology to fit and interpret first- and second-order regression models, and empirical process optimization methods derived from these. Also, issues related to efficient design of experiments to fit response surfaces.
Current Textbook: Empirical Model-Building and Response Surfaces, by George E.P. Box and Norman R. Draper, John Wiley and Sons, 1987.
Topics Covered | Chapters | Time |
Introduction; Matrix algebra review | 1,2, Appendix 3 C | 1 week |
Regression review | 3 | 1 week |
Graphical representations of functions | handouts | 0.5 weeks |
Fitting and interpreting first order surfaces | 4,5 | 2 weeks |
Fitting and interpreting second order surfaces | 7,9,10,11 | 3 weeks |
Transformations | 8 | 1 week |
Steepest ascent and other optimization methods | 6 | 1.5 weeks |
Specialized experimental designs for second order surfaces | 13,14 | 1.5 weeks |
Experiments with mixtures | handouts | 1 week |
Nonlinear regression models | 12, Appendix 3D | 1 week |
Contact Faculty: TBD