Table 1.

Metrics and their description.

Metric	Description
Accuracy	This is the number of currently classified samples in a validation set, divided by the total number of samples.
Balanced Accuracy	This is the average of the per-class accuracies in a validation set; per-class accuracy is one way to account for imbalance in the number of samples in each class.
Sensitivity	This is the same as true positive rate (TPR) or recall; it measures the fraction of a designated ‘positive class’ (e.g., ‘tumour’) that are correctly classified.
Specificity	This is the same as true negative rate (TNR); it measures the fraction of a designated ‘negative class’ (e.g., ‘normal’) that are correctly classified.
AUC	This is a measure of the quality of binary classifier based on the classifier’s confidence scores on the validation set; it is determined without regard to the selection of a single fixed threshold for separating classes.