Control Phase 3

A Black Belt course on advanced regression and advanced DOE. The first half fixes the regression problems flagged by residual diagnostics — outliers, non-linearity, and non-constant variance — then builds multiple-regression models and checks them for autocorrelation, collinearity, and best subsets. The second half covers response surface methodology: RSM foundations, steepest ascent/descent, EVOP, and central composite designs, finishing with a worked example that screens the region and builds the optimal surface.

What You'll Learn:

- Advanced regression — correcting failed residuals: outliers, non-linearity, non-constant variance

- Multiple regression and diagnostics: autocorrelation, collinearity, best subsets

- Advanced DOE — RSM foundations and response-surface types

- Steepest ascent/descent, EVOP (factorial and simplex), and central composite designs

- A worked RSM-CCD example: screen the region, build the surface, and optimize

Who This Is For:

Black Belt candidates and senior engineers who need to repair regression models and optimize processes with response surface methods.