Figure 4

AUPRC calculated for the statistical, XGBOOST and Transformer models. (a) combined performance of all classes per model. Precision is the measure of correctly identified positive cases from all the cases predicted as positive. Recall is the measure of correctly identified positive cases from all the actual positive cases. Precision-Recall curves which demonstrate the low false positive rate desired when precision is high and low false negative rate when recall is high are calculated for (b) Low class, (c) Medium class and (d) High class. (e) Per class AUPRC performance for the Transformer, XGBOOST and Statistical model.