Principle:Scikit learn contrib Imbalanced learn Sensitivity Specificity Analysis
| Knowledge Sources | |
|---|---|
| Domains | Machine_Learning, Model_Evaluation, Imbalanced_Learning |
| Last Updated | 2026-02-09 03:00 GMT |
Overview
A pair of complementary evaluation metrics that separately measure a classifier's ability to correctly identify positive cases (sensitivity) and negative cases (specificity), providing a complete picture of classification behavior on imbalanced data.
Description
Sensitivity (also called recall or true positive rate) measures the proportion of actual positives correctly identified: TP / (TP + FN). Specificity (true negative rate) measures the proportion of actual negatives correctly identified: TN / (TN + FP). Together, they reveal whether a classifier is biased toward one class.
Standard accuracy conflates these two aspects. Reporting them separately is essential for imbalanced classification where the cost of false negatives and false positives may differ significantly.
Usage
Use these metrics when you need to understand both positive and negative class recognition separately, especially in medical, fraud detection, or other domains where false negative/positive costs are asymmetric.
Theoretical Basis
The internal sensitivity_specificity_support function computes both values simultaneously along with per-class support counts.