Scikit-Learn GridSearchCV Lasso

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 Lasso regression, a popular algorithm for linear regression tasks with L1 regularization.

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.

Lasso regression is a linear model that incorporates L1 regularization to enforce sparsity in the model coefficients, making it useful for feature selection in high-dimensional datasets. The model is trained by minimizing the least squares loss function with an added penalty proportional to the sum of the absolute values of the coefficients.

The key hyperparameters for Lasso regression include the regularization strength (alpha), which controls the model complexity and helps prevent overfitting, and the max_iter, which sets the maximum number of iterations for the solver to converge.

from sklearn.datasets import make_regression
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.linear_model import Lasso

# Generate synthetic regression dataset
X, y = make_regression(n_samples=1000, n_features=20, n_informative=10, 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 = {
    'alpha': [0.01, 0.1, 1, 10],
    'max_iter': [1000, 5000, 10000]
}

# Perform grid search
grid_search = GridSearchCV(estimator=Lasso(random_state=42),
                           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 score: {test_score:.3f}")

Running the example gives an output like:

Best score: -0.011
Best parameters: {'alpha': 0.01, 'max_iter': 1000}
Test set score: 1.000

The steps are as follows:

1. Generate a synthetic regression dataset using scikit-learn’s make_regression function.
2. Split the dataset into train and test sets using train_test_split.
3. Define the parameter grid with different values for alpha and max_iter hyperparameters.
4. Perform grid search using GridSearchCV, specifying the Lasso 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 score.

By using GridSearchCV, we can easily explore different hyperparameter settings and find the combination that maximizes the model’s performance. This automated approach saves time and effort compared to manual hyperparameter tuning and helps ensure we select the best configuration for our Lasso regression model.