Mixture experiments with multiple response (Y's) to meet

nil

Hi JMP Community,

I am trying to understand how to proceed for using mixture OFAT experiment results (retrospective analysis ) to identify setting which meets requirement.

This is particular for Pharmaceutical table composition,which have 4 components varied and dissolution is performed in 3 different conditions, and 9 OFAT experiments data is available.

I am trying to arrive at composition which can meet required criteria is all 3 conditions.

 

Looking for help particularly on following aspects

1.  How to use OFAT mixture experiment data, to run mixture models with linear blending or non-linear blending using Fit Model (using JMP 15).

2. What are possible mixture models which can be fit with 9 experiments

3. How to analyze for combined consideration to meets multiple Y's, like meeting acceptance criteria/spec in 3 different conditions as discussed above.

 

Your help on this is appreciated.

 

Thanks!

 

 

Mark_Bailey

You have 9 runs after combining the OFAT experiments or 9 runs in each OFAT experiment?

 

It is unlikely that you can model non-linear blending with data from OFAT experiments.

 

In this situation, it is not just the amount of data but the nature of the data that determines what model may be estimated.

 

You can enter more than one response in the Fit Least Squares platform and jointly optimize them with the Prediction Profiler using individual desirability function for each response.