Table 1. Confusion Matrix, Accuracy, and other evaluation metrics obtained after combining 15 test MB datasets followed by applying the classifier on the combined dataset (N = 1,286 samples).
| Confusion Matrix | ||||
|---|---|---|---|---|
| Ref_Group3 | Ref_Group4 | Ref_WNT | Ref_SHH | |
| Pred_Group3 | 210 | 10 | 0 | 1 |
| Pred_Group4 | 5 | 478 | 0 | 0 |
| Pred_WNT | 0 | 0 | 110 | 2 |
| Pred_SHH | 3 | 7 | 0 | 392 |
| Overall Stats | ||||
| Accuracy | 97.70% | |||
| Kappa | 96.70% | |||
| AccuracyLower | 96.70% | |||
| AccuracyUpper | 98.50% | |||
| AccuracyNull | 40.60% | |||
| AccuracyPValue | 0.00E+00 | |||
| McnemarPValue | NaN | |||
| Class Stats | ||||
| Sensitivity | Specificity | Pos Pred Value | Neg Pred Value | |
| Class: Group3 | 96.30% | 98.90% | 95.00% | 99.20% |
| Class: Group4 | 96.60% | 99.30% | 99% | 97.70% |
| Class: WNT | 100% | 99.80% | 98.20% | 100% |
| Class: SHH | 99.20% | 98.80% | 97.50% | 99.60% |
| Precision | Recall | F1 | Prevalence | |
| Class: Group3 | 95.00% | 96.30% | 95.70% | 17.90% |
| Class: Group4 | 99% | 96.60% | 97.80% | 40.64% |
| Class: WNT | 98.20% | 100% | 99.10% | 9.03% |
| Class: SHH | 97.50% | 99.20% | 98.40% | 32.43% |
| Detection Rate | Detection Prevalence | Balanced Accuracy | ||
| Class: Group3 | 17.24% | 18.10% | 97.60% | |
| Class: Group4 | 39.24% | 39.70% | 98% | |
| Class: WNT | 9.03% | 9.20% | 99.90% | |
| Class: SHH | 32.18% | 33.00% | 99.00% |