I am using TMVA to train a classification task, but I am currently confused about the settings for NormNode and weight. Since the backgrounds I am using come from different energy regions in the MC files, I used AddBackgroundTree to add trees from different files, and set background weights according to the proportion of different energy ranges.
However, the spectrum in each MC file for the different energy regions does not match the actual spectrum shape, so I set a reweight variable for each tree and applied it using SetBackgroundWeightExpression(“reweight”) to achieve event-by-event reweighting.
With these settings, I can account for multiple energy ranges and correct the spectrum differences between the MC and data. However, I am unsure how to address potential sample imbalance issues caused by these settings (it is difficult to directly adjust the number of events or weights of multiple backgrounds to balance them with the signal). I noticed that NormNode might offer some help, such as NumEvents and EqualNumEvents, but I am uncertain whether they are effective for my specific case.
Please read tips for efficient and successful posting and posting code
Please fill also the fields below. Note that root -b -q will tell you this info, and starting from 6.28/06 upwards, you can call .forum bug from the ROOT prompt to pre-populate a topic.
ROOT Version: Not Provided
Platform: Not Provided
Compiler: Not Provided