The features & target in my dataset are very skewed. Could anyone explain why transforming the features & target (I'm using a Yeo-Johnson transformation) is significantly improving the performance of my model when using a decision tree?
Asked
Active
Viewed 19 times
