Fig. 5: Performance evaluation of the XGBoost classifier after CV. | npj Materials Degradation

Fig. 5: Performance evaluation of the XGBoost classifier after CV.

From: Predicting stress corrosion cracking in downhole environments: a Bayesian network approach for duplex stainless steels

Fig. 5

a Confusion matrix displaying the aggregated results from the CV process, where the colour bar indicates the number of instances evaluated across test sets. The XGBoost classifier correctly classified 1682 TP and 1248 TN, incurring 144 FP and 112 FN. b ROC curves for each CV fold. The AUC scores range from 0.899 to 0.991. The mean AUC score was 0.967, with a standard deviation (shaded in grey) of ± 0.036, indicating a high model’s discriminative ability across diverse dataset segments. The overall accuracy of the XGBoost model was 91.97%.

Back to article page