[gboost] extend tables weak learner to scalars

accosmin-org / libnano

C++ numerical optimization and machine learning utilities using Eigen3

MIT License

4 stars 1 forks source link

Open accosmin opened 2 years ago

accosmin commented 2 years ago

predict a constant by binning the scalar features using configurable quantiles (e.g. of 10% increments)
extend the unit tests accordingly