Model drift detection in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the library used for detecting model drift.

ML Python

from sklearn.metrics import [1]

Drag options to blanks, or click blank then click option'

Amean_squared_error

Baccuracy_score

Cclassification_report

Droc_auc_score

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing a metric that measures error instead of accuracy.

Importing a report function instead of a metric.

The accuracy_score function helps measure how well a model performs, which is useful for detecting model drift by comparing performance over time.

2fill in blank

medium

Complete the code to calculate the accuracy of model predictions.

ML Python

accuracy = accuracy_score(y_true, [1])

Drag options to blanks, or click blank then click option'

Ay_prob

By_test

Cy_train

Dy_pred

Attempts:

3 left

💡 Hint

Common Mistakes

Using predicted probabilities instead of predicted labels.

Confusing test data with predictions.

The accuracy_score function compares the true labels y_true with the predicted labels y_pred to calculate accuracy.

3fill in blank

hard

Fix the error in the code to detect model drift by comparing old and new accuracy.

ML Python

if new_accuracy [1] old_accuracy - threshold:
    print('Model drift detected')

Drag options to blanks, or click blank then click option'

A>

B==

C<

D>=

Attempts:

3 left

💡 Hint

Common Mistakes

Using greater than operator which means performance improved.

Using equality operator which doesn't detect drift.

Model drift is detected when the new accuracy is less than the old accuracy minus a threshold, indicating performance dropped.

4fill in blank

hard

Fill both blanks to create a dictionary that stores accuracy scores for old and new data.

ML Python

accuracy_scores = {'old': [1], 'new': [2]

Drag options to blanks, or click blank then click option'

Aold_accuracy

Bnew_accuracy

Caccuracy_score(y_old_true, y_old_pred)

Daccuracy_score(y_new_true, y_new_pred)

Attempts:

3 left

💡 Hint

Common Mistakes

Calling accuracy_score inside the dictionary instead of using variables.

Swapping old and new accuracy values.

The dictionary stores the pre-calculated accuracy values old_accuracy and new_accuracy for easy comparison.

5fill in blank

hard

Fill all three blanks to complete the code that calculates accuracy for old and new data and checks for drift.

ML Python

old_accuracy = [1](y_old_true, y_old_pred)
new_accuracy = [2](y_new_true, y_new_pred)
if new_accuracy [3] old_accuracy - 0.05:
    print('Model drift detected')

Drag options to blanks, or click blank then click option'

Aaccuracy_score

C<

D>

Attempts:

3 left

💡 Hint

Common Mistakes

Using different functions for accuracy calculation.

Using greater than operator which means performance improved.

Both accuracies are calculated using accuracy_score. The condition checks if new accuracy dropped below old accuracy minus 0.05, signaling drift.