I have done the medium optimization using RSM - CCD for two factors (Carbon source (A) & Nitrogen source (B)) with the levels ranging from 137-222 for A, 3-15 for B. As a result am getting optimal level for B at 12. But for A I am getting too far value 222 as optima. The surface plot is following an increasing trend for A, center point for factor A is out of the design. Increasing the A value corresponding response (Product titer) value also increasing proportionally.
I want to know weather this model is acceptable or not. If not please let me know what to do to get the best optimal design.
That is fairly common. What is happening is your software is using the regression model you generated and trying to find the Optimal point (Max or Min). Even though you did not test a points at this combination, the regression model seems to think it will be the best.
I think #Andrew Ekstrom is right that model is following linear regression and it seems that the region selection may be causing problem. I can suggest try for the good region selection so that you may achieve a good model with appropriate results.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics