Table 3 Confusion matrix displaying the relation between different targets and the ensemble estimate
From: Neural network analysis of sleep stages enables efficient diagnosis of narcolepsy
Target | ||||||
|---|---|---|---|---|---|---|
Model Predictions | Wake | N1 | N2 | N3 | REM | Precision |
Wake | 14.08% | 0.35% | 0.88% | 0.007% | 0.08% | 0.91 |
16.68% | 0.15% | 0.44% | 0.003% | 0.02% | 0.96 | |
N1 | 1.13% | 1.78% | 3.00% | 0.002% | 0.36% | 0.28 |
0.47% | 0.88% | 1.15% | 0% | 0.12% | 0.34 | |
N2 | 0.29% | 0.59% | 52.58% | 1.27% | 0.66% | 0.95 |
0.12% | 0.25% | 56.30% | 0.34% | 0.32% | 0.98 | |
N3 | 0.002% | 0% | 2.13% | 4.87% | 0% | 0.70 |
0% | 0% | 1.09% | 4.23% | 0% | 0.91 | |
REM | 0.54% | 1.17% | 0.78% | 0% | 13.45% | 0.84 |
0.40% | 0.73% | 0.41% | 0% | 15.86% | 0.91 | |
Sensitivity | 0.88 | 0.46 | 0.89 | 0.79 | 0.92 | 0.87 |
0.94 | 0.44 | 0.95 | 0.92 | 0.97 | 0.94 | |
