Scikit-Learn GridSearchCV BaggingRegressor

Hyperparameter tuning is a crucial step in optimizing machine learning models for best performance. In this example, we’ll demonstrate how to use scikit-learn’s GridSearchCV to perform hyperparameter tuning for BaggingRegressor, an ensemble algorithm used for regression tasks.

Grid search is a method for evaluating different combinations of model hyperparameters to find the best performing configuration. It exhaustively searches through a specified parameter grid, trains and evaluates the model for each combination using cross-validation, and selects the hyperparameters that yield the best performance metric.

BaggingRegressor is an ensemble algorithm that improves the stability and accuracy of machine learning models by averaging multiple versions of a predictor, reducing variance and preventing overfitting. The model is particularly useful when dealing with high variance, low bias models like decision trees.

The key hyperparameters for BaggingRegressor include the number of base estimators (n_estimators), which determines how many trees will be used in the ensemble; max_samples, which controls the number of samples to draw from the training set for each base estimator; and max_features, which specifies the number of features to draw from the training set for each base estimator.

from sklearn.datasets import make_regression
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.ensemble import BaggingRegressor
from sklearn.tree import DecisionTreeRegressor

# Generate synthetic regression dataset
X, y = make_regression(n_samples=1000, n_features=10, n_informative=8, noise=0.1, random_state=42)

# Split into train and test sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Define parameter grid
param_grid = {
    'n_estimators': [10, 50, 100],
    'max_samples': [0.5, 1.0],
    'max_features': [0.5, 1.0]
}

# Perform grid search
grid_search = GridSearchCV(estimator=BaggingRegressor(estimator=DecisionTreeRegressor()),
                           param_grid=param_grid,
                           cv=5,
                           scoring='neg_mean_squared_error')
grid_search.fit(X_train, y_train)

# Report best score and parameters
print(f"Best score: {grid_search.best_score_:.3f}")
print(f"Best parameters: {grid_search.best_params_}")

# Evaluate on test set
best_model = grid_search.best_estimator_
test_score = best_model.score(X_test, y_test)
print(f"Test set R^2: {test_score:.3f}")

Running the example gives an output like:

Best score: -1303.000
Best parameters: {'max_features': 1.0, 'max_samples': 1.0, 'n_estimators': 100}
Test set R^2: 0.895

The steps are as follows:

1. Generate a synthetic regression dataset using make_regression.
2. Split the dataset into train and test sets using train_test_split.
3. Define the parameter grid with different values for n_estimators, max_samples, and max_features hyperparameters.
4. Perform grid search using GridSearchCV, specifying the BaggingRegressor model, parameter grid, 5-fold cross-validation, and negative mean squared error as the scoring metric.
5. Report the best cross-validation score and best set of hyperparameters found by grid search.
6. Evaluate the best model on the hold-out test set and report the R-squared score.

By using GridSearchCV, we can efficiently explore different hyperparameter settings and find the configuration that maximizes the performance of the BaggingRegressor model. This automated approach streamlines the process of hyperparameter tuning, ensuring optimal model performance.