Calculate sensitivity, the true positive rate, from a confusion matrix.
Sensitivity, also called the true positive rate (TPR) or recall, measures how many of the actual positive cases the model correctly caught: sensitivity = TP / (TP + FN). It's mathematically identical to recall — the different name is conventional in medical and diagnostic testing contexts, while 'recall' is more common in machine learning and information retrieval. Sensitivity and specificity together form the two axes of the ROC curve and fully characterize a binary classifier's tradeoffs.
sensitivity = TP / (TP + FN)
Yes, sensitivity and recall are computed identically as TP / (TP + FN); the terminology difference is purely conventional across fields — 'sensitivity' in medicine/diagnostics, 'recall' in ML/information retrieval.
100% sensitivity means the test/model correctly identified every single true positive case, with zero false negatives — though this alone says nothing about how many false positives it also produced.
Adjusting a model's decision threshold to increase sensitivity (catching more true positives) typically decreases specificity (more false alarms among negatives), and vice versa — the ROC curve visualizes this tradeoff across all thresholds.