If I use IterativeImputer with the estimator “RandomForestRegressor()” and, on the other side, MissForest Imputer, what is the difference ?
Iterative imputer will use tree-based methods to decide which value is correct for imputation (like RandomForestRegressor), but this will be the same as for Missforest which also uses tree-based method (also a random forest). In my understanding, both are equivalent.
But my guess is that both are different, somehow (and not only on their potential capacity in dealing with numerical/categorical data). Can you explain their “real” differences? Thank you.