Understanding XGBoost

Hey I’m trying to create a Federated Learning implementation of XGBoost in python so that I can play around with the different parameters. I am not experienced with big projects like these and am having difficulty finding the c++ files in which XGBoost is implemented.

  • Specifically I’d like to know how multi-class classification works in XGBoost, where could I find that in the code?
  • And how is the first tree constructed, and where would I be able to see that in the code?

Thanks in advance if you know where this code would be!

1 Like