Model Training & Hyperparameter Tuning

5/8

Model Selection Strategy

Random Forest

R² = 0.81

n_estimators: [8, 16, 32, 64, 128, 256]

Gradient Boosting

R² = 0.83

learning_rate, subsample, n_estimators

XGBoost

R² = 0.84

learning_rate, n_estimators

CatBoost Regressor

R² = 0.87

depth, learning_rate, iterations

Hyperparameter Tuning

GridSearchCV with 5-fold cross-validation

Evaluation metric: R² Score

Model Performance

Best Model Features

  • High accuracy on test data
  • Handles categorical features
  • Robust to outliers

Evaluation Metrics

R² Score: 0.87
MAE: 3.42
RMSE: 4.61
Prev Slide 5/8 Next
Prev Slide 5/8 Next