KRR model implementation on the diabetes dataset

krr_code.py

from sklearn.datasets import load_diabetes
from sklearn.kernel_ridge import KernelRidge
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.metrics import mean_absolute_error, mean_squared_error

diabetes = load_diabetes()
data = diabetes.data
target = diabetes.target

X_train, X_test, y_train, y_test = train_test_split(data, target, test_size=0.2)

krr_model = KernelRidge()
param_grid = {
    "alpha": [1e-5, 0.00001, 0.0001, 0.001, 0.01, 0.1, 1, 2],
    "kernel": [
        "linear",
        "rbf",
        "poly",
        "sigmoid",
    ],
}
grid_search = GridSearchCV(
    krr_model, param_grid, scoring="neg_mean_absolute_error", n_jobs=-1, cv=5
)

grid_search.fit(X_train, y_train)
best_params = grid_search.best_params_
print(best_params)
best_model = grid_search.best_estimator_

predictions = best_model.predict(X_test)
test_mae = mean_absolute_error(y_test, predictions)
test_mse = mean_squared_error(y_test, predictions)

train_predictions = best_model.predict(X_train)
train_mae = mean_absolute_error(y_train, train_predictions)
train_mse = mean_squared_error(y_train, train_predictions)

print(f"Test MAE : {test_mae} and Test MSE : {test_mse}")
print(f"Train MAE : {train_mae} and train MSE : {train_mse}")