5 ms·
I would say what we really care about is Sensitivity and Specificity rather than a confusion matrix, given the confusion matrix is a function of the actual posi
by hamburglar1 6y ago
I would say what we really care about is Sensitivity and Specificity rather than a confusion matrix, given the confusion matrix is a function of the actual positive incidence rate of the population tested and consequently a problematic measure in this case.
- hprotagonist 6y agowell, how do you get one without the other? either way i haven’t seen ROC, precision/recall, d’, or F1 scores either.