3

I'm trying to build a random forest with 100k records and 2K variables. I have an imputation process to handle missing values while using randomForest but I want to understand what exactly happens to missing values in predictors in cforest.

I looked at a tree and there were no specific actions for missing values. The code says na.action = NULL. Is there any imputation going on internally in cforest or it can handle NAs as separate entity? Also is it advisable to use cforest instead of randomForest to deal with missing values?

Sourav Sarkar
  • 31
  • 1
  • [Please don't cross-post](http://stackoverflow.com/questions/33008784/missing-value-handling-in-cforestparty-in-r) –  Oct 08 '15 at 07:13

0 Answers0