Details
-
Improvement
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
-
None
Description
Missing value handling in RandomForest is missing in the current implementation.
So, missing value (null/NaN) should be handled efficiently in RandomForest.
http://oro.open.ac.uk/22531/1/decision_trees.pdf
https://www.slideshare.net/kato_kohaku/imputation-of-missing-values-using-random-forest
Attachments
Issue Links
- is related to
-
HIVEMALL-163 Add IS_INFINITE, IS_FINITE, IS_NAN functions
- Closed