/XGBoost-tuning

Primary LanguageJupyter Notebook

XGBoost tuning using Random Search

This repository contains an example of XGBoost hypeparameter tuning using Random Search.

  • Provides a function capable to tune flexibly XGBoost for K-fold cross validation and early stopping.
  • Results of tuning can be easily manipulated using number of estimators.
  • Results of tuning can be well evaluated seeing the error on training, validation and test sets.