Fairness metrics in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to calculate the accuracy of a model.

ML Python

accuracy = correct_predictions / [1]

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

Apredictions

Bcorrect_predictions

Ctotal_predictions

Dtotal_labels

Attempts:

3 left

💡 Hint

Common Mistakes

Dividing by correct_predictions instead of total_predictions.

Using predictions instead of total count.

Accuracy is the number of correct predictions divided by the total number of predictions made.

2fill in blank

medium

Complete the code to calculate demographic parity difference between two groups.

ML Python

dp_difference = abs(rate_group1 - [1])

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

Arate_group1

Brate_group2

Ctotal_group2

Dtotal_group1

Attempts:

3 left

💡 Hint

Common Mistakes

Subtracting the same group's rate.

Using total counts instead of rates.

Demographic parity difference is the absolute difference between positive rates of two groups.

3fill in blank

hard

Fix the error in the code to compute equal opportunity difference.

ML Python

equal_opportunity_diff = tpr_group1 - [1]

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

Atpr_group2

Btpr_group1

Cfpr_group2

Dfpr_group1

Attempts:

3 left

💡 Hint

Common Mistakes

Using false positive rates instead of true positive rates.

Subtracting the same group's TPR.

Equal opportunity difference compares true positive rates (TPR) between groups.

4fill in blank

hard

Fill both blanks to calculate disparate impact ratio.

ML Python

disparate_impact = [1] / [2]

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

Apositive_rate_minority

Bpositive_rate_majority

Ctotal_minority

Dtotal_majority

Attempts:

3 left

💡 Hint

Common Mistakes

Using total counts instead of positive rates.

Reversing numerator and denominator.

Disparate impact is the ratio of positive rates between minority and majority groups.

5fill in blank

hard

Fill all three blanks to create a fairness report dictionary.

ML Python

fairness_report = {
    'demographic_parity_diff': [1],
    'equal_opportunity_diff': [2],
    'disparate_impact': [3]
}

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

Adp_difference

Bequal_opportunity_diff

Cdisparate_impact

Daccuracy

Attempts:

3 left

💡 Hint

Common Mistakes

Including accuracy instead of fairness metrics.

Mixing up metric variable names.

The fairness report includes demographic parity difference, equal opportunity difference, and disparate impact metrics.