Table 11.
Performance results for the multi-layer perceptron algorithm on each of the outer cross-validation test sets, for the male sample.
Model | MAEa (years) | Accuracy | AUCb | RMSEc (years) | Recall | Precision |
1 (median) | 0.95 | 0.33 | 0.83 | 1.29 | 0.91 | 0.48 |
2 | 1.08 | 0.30 | 0.85 | 1.40 | 073 | 0.35 |
3 | 0.89 | 0.32 | 0.84 | 1.17 | 0.17 | 1.00 |
4 | 0.91 | 0.33 | 0.83 | 1.23 | 0.83 | 0.59 |
5 | 1.05 | 0.35 | 0.84 | 1.49 | 0.42 | 0.83 |
aMAE: mean absolute error.
bAUC: area under the curve.
cRMSE: root mean squared error.