The
model comparison tool enables you to compare the performance of
competing models using various benchmarking criteria. The comparison criteria available depends on
the models and
response variable used in your analysis. A model comparison requires that at least one model is trained
before you can perform a comparison.
Before performing a model comparison, ensure that all models are initialized and updated.
If the
Auto-update model property is disabled
for a model, you must manually update it before you can compare it
to another model. A model is not considered initialized until it has
been trained.
When you change a model after a comparison has been created, changes are not carried
over to the model comparison.