Fig. 7.

Rate of total classification accuracy for all classes, and true positive rate (TPR, same as sensitivity) and true negative rate (TNR, same as specificity) for bleeding, as a function of number of samples per class, for the simulated data. Total accuracy was calculated by summing all correctly classified observations (the main diagonal of the confusion matrix, see Table 2 that shows the result for 250 samples per class) and dividing by the total number of observations. To calculate TPR and TNR for bleeding, all bleeding sizes were combined into one class