is xgboost classifier affected by distribution of features? my thinking is it is not , e.g. if data is skewed it does not need to be made more normally distributed. am i right?
Distribution of variables
No, XGBoost does not make any assumption about how features are distributed.
1 Like