Figure A3
The chart is titled “ROC Curves – All Models (Fatal H R I s equals Positive Class)”. The horizontal axis is labeled “False Positive Rate”, ranging from 0.0 to 1.0 with an interval of 0.2, and the vertical axis is labeled “True Positive Rate”, also ranging from 0.0 to 1.0 with an interval of 0.2. A diagonal dashed line from the bottom left at (0.0, 0.0) to the top right at (1.0, 1.0) represents a random classifier with an A U C of 0.5. Four colored curves represent different machine learning models. The orange curve for “Random Forest” has an A U C of 0.772 and generally lies above the others across most of the range. The green curve for “X G Boost” has an A U C of 0.762 and follows closely behind Random Forest. The purple curve for “Logistic Regression” has an A U C of 0.766 and performs similarly to the top models, overlapping in several regions. The blue curve for “K N N” has a lower A U C of 0.680 and remains below the other model curves for most values of false positive rate. All curves start near the origin and rise toward the top right corner, with improved true positive rates as false positive rates increase. The Random Forest, Logistic Regression, and X G Boost models show better separation from the diagonal baseline, indicating stronger predictive performance compared to K N N. Note: All numerical data values are approximated.The ROC curves of the classification models. Source: Authors’ own work