Table 3 Performance Metrics Across ASD Datasets: Summary of Precision, Recall, F1 Score, Accuracy, ROC-AUC, and P-value for Logistic Regression, Random Forest, and TabNet models evaluated using 5-Fold Stratified Cross-Validation on the Nadig, Eigsti, Eigsti Multi, and Merged datasets. Metrics were calculated with standard formulas and validated through cross-validation to ensure reliability. Bold values indicate the highest score for each metric within the corresponding dataset.
From: Screening autism spectrum disorder in children using machine learning on speech transcripts
Datasets | Metric | Logistic regression | Random forest | TabNet |
|---|---|---|---|---|
Nadig | Precision | 0.933 | 0.833 | 0.933 |
Recall | 0.933 | 0.767 | 0.867 | |
F1 | 0.920 | 0.760 | 0.860 | |
Accuracy | 0.946 | 0.864 | 0.921 | |
ROC-AUC | 0.933 | 0.927 | 0.860 | |
P-value | 0.00017 | 0.0037 | 0.0011 | |
Eigsti | Precision | 0.867 | 0.800 | 0.760 |
Recall | 0.767 | 0.750 | 0.933 | |
F1 | 0.807 | 0.729 | 0.835 | |
Accuracy | 0.824 | 0.748 | 0.819 | |
ROC-AUC | 0.878 | 0.817 | 0.889 | |
P-value | 0.017 | 0.020 | 0.009 | |
Eigsti Multi | Precision | 0.692 | 0.802 | 0.647 |
Recall | 0.577 | 0.753 | 0.588 | |
F1 (weighted) | 0.584 | 0.763 | 0.596 | |
Accuracy | 0.577 | 0.753 | 0.588 | |
ROC-AUC | 0.553 | 0.710 | 0.578 | |
P-value | 0.121 | 0.002 | 0.258 | |
Merged | Precision | 0.869 | 0.844 | 0.880 |
Recall | 0.857 | 0.827 | 0.869 | |
F1 | 0.856 | 0.823 | 0.869 | |
Accuracy | 0.857 | 0.827 | 0.869 | |
ROC-AUC | 0.946 | 0.860 | 0.963 | |
P-value | 0.0013 | 0.003 | 0.0014 |
