Average Odds Ratio

Overview

The Average Odds Ratio is a fairness metric used to assess the ratio in predictive performance between two groups or populations in terms of both false positive rates and true positive rates. It focuses on measuring the balance of prediction outcomes across different groups.

Formula

Average Odds Ratio = (ratio between groups' true positive rates + ratio between groups' true negative rates) / 2

Where:

True Positive Rate = True Positives / (True Positives + False Negatives)
False Negative Rate = False Negatives / (True Positives + False Negatives)

Source

Usage

Manually

# Calculate average odds ratio
result = average_odds_ratio(df, protected_attribute, privileged_group, labels, positive_label, y_true)

print("Average Odds Ratio:", result)

Using Fairness Object

result = (fo.compute(average_odds_ratio))

Results

Average Odds Ratio: 1.249999999653125

These results are obtained by using the input data given in the Create Example Data page under Getting Started

Interpretation

A positive Average Odds Ratio indicates that the odds of false positives and true positives are higher for the first group compared to the second group. This suggests a potential imbalance in predictive performance, where the model may be more likely to classify members of the first group incorrectly and members of the second group correctly.

Conversely, a negative Average Odds Ratio suggests that the odds of false positives and true positives are higher for the second group. This indicates a potential disparity in predictive performance, where the model may be more likely to misclassify members of the second group while correctly classifying members of the first group.