Contains the tools that mainly developed for VHH analysis on CMS experiment. But they are flexible!
BDT classifiers based on XGBoost with the python sk-learn interface/API. They can be used for variable plotting, correlation calculation, etc. Also the training results are also visiable, such as variable importances, validation plots, etc.
Now the main parts of this tool has been independantly implemented in BDT4HEP, but you can still find examples in this part.
The implemented scripts to generate samples from the truth level until NANO-AOD files.
The scripts in this part can be used for simple tasks that are useful for Higgs pair studies. For example, to calculate the interference between the Feynman diagrams in ggHH/VBFHH/VHH production mode. There are also scripts for ROOT based analysis tasks such as a self defined tree slimmer. What's more, for official tools recommended by CMS higgs group, the HiggsCombinationTool and HH inference tool, you can find some usefull commands or scripts in HiggsCombine. More details and usage can be found at this note.
Boost Decision Tree is one of the most popular Reinforcement Learning technique adapted in Particle Physics experiment study. This repo provide the python interface to apply BDT to the HEP data.
TMVA is a machine learning interface in ROOT. This repo give an example about doing classification with TMVA with pyROOT. It is a very easy using and adapting tool for beginner. And a GUI micro based on ROOT can also be found to do the visialization and evaluation of the BDT training.