In this step we will fine tune our models using cross validation. It is a resampling technique that is used to evaluate machine learning models on a limited data sample.
A test set should still be kept aside for final evaluation. We would no longer need a validation set (which is sometimes called the dev set) while doing cross validation. The training set is split into k smaller sets (there are other approaches too, but they generally follow the same principles). The following procedure is followed for each of the k folds:
k-1 of the folds as training datak-fold cross-validation is then the average of the values computed in the loop.Given below is a visual representation of this process:
This image is from the official page of the Scikit-learn cross validation where you can find more details about the process.
Now let's work on fine tuning our models using cross validation.
First, let's define a function called display_scores that would display the scores, mean, and standard deviation of all the models after applying cross validation
def <<your code goes here>>(scores):
print("Scores:", scores)
print("Mean:", scores.mean())
print("Standard deviation:", scores.std())
Now let's import cross_val_score from Scikit-learn
from sklearn.model_selection import <<your code goes here>>
Now let's calculate the cross validation scores for our Decision Tree model
scores = cross_val_score(tree_reg, housing_prepared, housing_labels,
scoring="neg_mean_squared_error", cv=10)
tree_rmse_scores = np.sqrt(-scores)
display_scores(tree_rmse_scores)
Finally, let's calculate the cross validation scores for our Random Forest model
forest_scores = cross_val_score(forest_reg, housing_prepared, housing_labels,
scoring="neg_mean_squared_error", cv=10)
forest_rmse_scores = np.sqrt(-forest_scores)
display_scores(forest_rmse_scores)
Note - Having trouble with the assessment engine? Follow the steps listed here
Loading comments...