Calculate BDT score with python based xgboost BDT

TMVA
1

Hi All,
I am trying to use python-based xgboost for my BDT analysis.

How do I calculate the BDT score after I am able to use “model.predict_proba”? I am looking for a snippet or a mathematical definition of BDT score after this step.

Thanks

2

Hi,

This is more a question for xgboost than ROOT. For ROOT TMVA you can look at the TMVA Users guide, in particular the section about decision trees.
See https://root.cern.ch/download/doc/tmva/TMVAUsersGuide.pdf

Lorenzo