Description
If the input features contain oversampled features from the Prepare Data for Predictions tool and the Training Data Excluded for Validation parameter value is not 0, the tool will set aside validation data with oversampled features. Features used in training could also be used to evaluate the model and this will lead to a very biased evaluation.
Solution
Set the Training Data Excluded for Validation parameter value to 0, or use data that was not oversampled in the Prepare Data for Predictions tool as the input features.