So, I am preparing a dataset for an ML algorithm, but I have run into a problem - the thing is that around 23 of 96 predictors have got IV more than 0.5 (the lowest is 1.7) and I am curious if it is really bad and could it cause overfitting of the model? Is there any way to approve/reject the hypothesis suggested by high IV of a predictor that it may cause overfitting?
Asked
Active
Viewed 3 times
