Boosted Decision Trees : New Plots

BDT was evaluated with the following setup:

NTrees=100:nEventsMin=20:MaxDepth=1:BoostType:AdaBoost:AdaBoostBeta=0.3:SeparationType:GiniIndex:nCuts=20

:PruneMethod=NoPruning

(a) Overtraining Plot

 

(b) Background rejection vs signal efficiency plot

 

(c) Efficiency Plot

 

(d) BDT Output

 

Classifier #signal,#background Optimal-cut S/sqrt(S+B) Nsig Nbkg EffSig EffBkg
BDT 3716,23714 -0.0992 36.77 2436 1953 0.656 0.082

 

Note : Please see the attached files for input variables and correlation matrix plots