Visit with Us - Wellness Tourism Package purchase model

Best model after hyperparameter tuning: xgboost

Best parameters: {'xgbclassifier__learning_rate': 0.1, 'xgbclassifier__max_depth': 5, 'xgbclassifier__n_estimators': 200}

Serving threshold: 0.45 (selected from out-of-fold training F1).

Test set metrics at the selected threshold

  • accuracy: 0.9091
  • precision: 0.7470
  • recall: 0.8000
  • f1: 0.7726
  • roc_auc: 0.9276

Trained on Marco8000/visit-with-us-dataset (train.csv / test.csv). Load with joblib.load('best_tourism_model_v1.joblib').

Downloads last month
-
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support