Previous Page

|

Next Page

Optimizing a Response

The goals of this experiment are to determine the levels of the control factors that bring MEAN(FORCE) to target and to minimize variability around that target. A high SNRL indicates small variability, so maximizing SNRL is part of the goal. With the desirability function in the prediction profiler, you can maximize SNRL at the same time that you optimize the control factor settings. Follow these steps:

1. Click Optimize.
2. Since ADX has not selected any significant effects for either FORCE or SNRL(FORCE), it uses the master model to optimize responses. Select FORCE and SNRL(FORCE) and click Run.
   
   

   
   
   

3. Select View Show Desirability to view desirability functions and traces.
4. Define desirability functions to maximize MEAN(FORCE) and SNRL(FORCE).
5. Move the factor settings to maximize overall desirability. Alternatively, you can select Settings Optimize desirability.