PGAE-ICA

Sensitivity (Se) is the proportion of correctly identified actual abnormal intelligence C&A by the models. Specificity (Sp) is the proportion of correctly identified normal intelligence C&A by the models. Positive predictive value (PV+) is the proportion of correctly identified C&A predicted to have abnormal intelligence. Negative predictive value (PV–) is the proportion of correctly identified C&A predicted to have normal intelligence. Odds ratio (OR) represents the ability of the models to distinguish between C&A with normal and abnormal intelligence. A larger OR indicates a higher ability to distinguish between abnormal and normal intelligence. Youden’s index (YI) represents the models’ ability to differentiate between normal and abnormal intelligence, assuming equal costs for false positives and false negatives. A larger YI reflects a greater ability to distinguish between abnormal and normal intelligence. The area under the curve (AUC) reflects the models’ classification performance. An AUC near 1 signifies superior classification performance.