I completed a four factor fractional factorial DOE with two different types of models, and I would like to use the prediction profiler and the simulator to help me optimize factors, accounting for their respective tolerance ranges. I am having trouble. The first response is a life variable. We have a mechanical wear type of phenomenon, and a Weibull distribution fits reasonably well. I used the parametric survival fit platform to create this model. The second response is nominal response variable, and so I have a nominal logistic regression model for that. With the separate model types, I believe I have to save the model formulas to the tables, then access the prediction profiler to reference each formula directly. This is instead of accessing the prediction profiler through the fit model platform. My problem is, when accessing the prediction profiler this way, the simulator doesn't seem to be an option. Am I missing something?
... View more