Regression Modeling Chapter 10 Design Analysis of Experiments
- Slides: 50
Regression Modeling Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 1
Design and Analysis of Engineering Experiments Ali Ahmad, Ph. D Chapter 10 Based on Design & Analysis of Experiments 7 E 2009 2
Regression models are empirical as opposed to mechanistic models Regression modeling is often performed on undesigned or unplanned data Regression modeling is also used extensively to build models to data from designed experiments Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 3
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 4
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 5
L is minimized by taking derivatives with respect to the model parameters and equating to zero Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 6
Taking the derivatives and equating to zero results in: Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 7
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 8
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 9
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 10
Mean square error Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 11
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 12
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 13
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 14
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 15
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 16
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 17
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 18
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 19
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 20
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 21
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 22
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 23
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 24
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 25
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 26
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 27
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 28
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 29
The test uses an ANOVA approach. The regression or model sum of squares is See Table 10. 4 for the viscosity regression model Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 30
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 31
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 32
See Table 10. 4 for the viscosity regression model Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 33
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 34
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 35
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 36
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 37
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 38
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 39
Prediction intervals are useful when running conformation experiments If the new observation falls within the prediction interval that is some evidence that the model is reliable Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 40
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 41
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 42
10. 7 Regression Model Diagnostics Scaled residuals and PRESS Standardized residual Studentized residual Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 43
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 44
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 45
Influence Diagnostics Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 46
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 47
10. 8 Testing for Lack of Fit Very important test in both general regression modeling and in analysis of a designed experiment Does the chosen model adequately fit the data, or should higherorder terms be considered? A statistical test can be performed provided that the error term can be decomposed into the two components shown below: Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 48
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 49
Chapter 10 Design & Analysis of Experiments 7 E 2009 Montgomery 50