Hi mate,
I wonder if there is common ways to perform back-transform predicted values from random forest model results pls?
I have run different random forest models for my data based on the transformed observed and predicted variables. Now I need to perform back-transform to the untransformed observed and predicted variables before re-calculating the error. I am not sure if back-transforming the error directly will work given the scaling issue of the transformations pls?
Thank you very much with kind regards
Bang Tran