Hello,
I'm developing a custom prediction model using AI Builder.
My dataset is unbalanced, the label is binary and only 5% of the records contain the positive class (the one I want to predict).
How can I train a model like this using AI Builder?
Oversampling? Generating syntetic data?