Model recommendations are only generated when you run full Autopilot. One of them
(the most accurate individual, non-blender model) will be prepared for deployment.
In the preparation process, DataRobot will: (1) calculate feature impact for the selected
model and use it to generate a reduced feature list, (2) retrain the selected model on the
reduced featurelist, (3) will replace the recommended model with the new model if
performance is improved on the reduced featurelist, (4) will retrain the model on a higher
sample size, and (5) will replace the recommended model with the higher sample size model if
it is more accurate.