README.md

xgboost-examples

XGBoost tutorial and examples for beginners

Basic

• Regression Hello World (Use XGBoost to fit xx curve)
• Classification Hello World (Use XGBoost to classify Breast Cancer Dataset)
• Fill Missing Values (Use Imputer to fill missing data)
• K-fold Cross Validation (Use K-fold to validate your model)
• Stratified K-fold CV (Use Stratified K-fold to make your split balanced)
• Visualize Single Tree (Plot single decision tree)

Intermediate

• Save And Reload Model - Approach 1 (Use pickle to save and reload your model)
• Save And Reload Model - Approach 2 (Use joblib to save and reload your model)
• Feature Importances (Two methods to visualize feature importances)
• Feature Selection (Use feature importances to select features)
• Early Stopping (Use Early Stopping to avoid overfitting)
• Plot Learning Curve (Use Learning Curve to judge your model)

Advanced

• Tune Number of Trees (Tune n_estimators parameter)
• Tune Size of Tree (Tune max_depth parameter)
• Tune Shrinkage (Tune learning_rate parameter)
• Tune Row Subsampling (Tune subsample parameter)
• Tune Column Subsampling by Tree (Tune colsample_bytree parameter)
• Tune Column Subsampling by Level (Tune colsample_bylevel parameter)