I see that in ML workflow, the oversampling of the data frame is performed after splitting the original data frame (into X and y), followed scaling and splitting for model test /train. I think, we should first perform oversampling (via ADASYN) in the original data frame and then perform the other ML workflow steps. Please review and clarify
I see that in ML workflow, the oversampling of the data frame is performed after splitting the original data frame (into X and y), followed scaling and splitting for model test /train. I think, we should first perform oversampling (via ADASYN) in the original data frame and then perform the other ML workflow steps. Please review and clarify