I have a small dataset of approximately 7000 rows, which is imbalanced. To balance the dataset I over and under sampled using the ROSE package. I then set up a probability m