What inspired this wish list request?
As a data scientist I have way too many requests to build scikit learn models so I would like regular users to be able to build their own models without having to learn how to code in python. Also the JMP Prediction Profiler is awesome and I would like to be able to use it to present model results to the people who asked me to build the model.
What is the improvement you would like to see?
I would like to transfer an sklearn prediction formula to JMP so it could be used in the prediction profiler because the prediction profiler runs so fast. I could call model.predict() to get the profiler values but that would be slow because the data would have to go back and forth from python to JMP. I am using the python environment within JMP to run the sklearn model.
Why is this idea important?
This would be so valuable because explaining and presenting complex machine learning models to non statisticians can be difficult. The JMP Profiler makes that much easier.