Sensitivity

Related: Specificity, Precision

$$\text{Sensitivity}=\mathrm{Pr}(\text{+ test | +disease})$$

• The sensitivity is the proportion of patients with the event that are classified as high risk

$$\text{Sensitivity}=\text{Recall}=\text{TPR}=\frac{TP}{TP+FN}$$

• Simple description: “how often is the prediction correct when positive value is predicted?”
• Worst value = 0; Best value = 1
• The sensitivity (aka true positive rate) is defined as $p(\text{test result}=1|\text{actual event}=1)$, i.e., the probability of a positive test given that the truth is positive.
• Sensitivity/Recall measures completeness—how many actual positives were predicted positives
• The False Negative Rate is defined as one minus the sensitivity, i.e. $\text{FNR}=1-\text{Sensitivity}$.