A Data Science Central Community
Just wanted to check if anyone has worked on gradient boosting technique (preferably using SAS E-miner) .
Great if you can share some good material/reference for the same.
- Introduction to gradient boosting
- Using gradient boosting in E-miner (which parameters to tweak etc.)
- Interpreting the output
If you are modeling rare events, tweak the leaf size to a smaller value. The default 0.1 might be too large for that case.
I second Eddy, press F1 to bring up EM Reference Help and see the doc for each node.
There should be some info in the sas online data mining community as well.