Using random forest for regression

The random forest algorithm is going to be the first ensemble to deal with here. It's an easy-to-grasp algorithm with straightforward hyperparameters. Nevertheless, as we usually do, we will start by training the algorithm using its default values, as follows, then explain its hyperparameters after that:

from sklearn.ensemble import RandomForestRegressor
rgr = RandomForestRegressor(n_jobs=-1)
rgr.fit(x_train, y_train)
y_test_pred = rgr.predict(x_test)

Since each tree is independent of the others, I set n_jobs to -1 to use my multiple processors to train the trees in parallel. Once they are trained and the predictions are obtained, we can print the following accuracy metrics:

from sklearn.metrics import (
    mean_squared_error, mean_absolute_error, median_absolute_error, r2_score
)

print(
    'R2: {:.2f}, MSE: {:.2f}, RMSE: {:.2f}, MAE {:.2f}'.format(
        r2_score(y_test, y_test_pred),
        mean_squared_error...