TABLE 1. Results.
| Dataset | Partition | Specificity | Sensitivity | Precision | Accuracy | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Chance | Original | LLE | PCA | Chance | Original | LLE | PCA | Chance | Original | LLE | PCA | Chance | Original | LLE | PCA | ||
| Beijing (Peking_3) | Training | 100% | 73.3% ± 15.8% | 80% ± 14.3% | 66.7% ± 16.9% | 0% | 53.3% ± 17.9% | 93.3% ± 8.9% | 80% ± 14.3% | 0% | 66.7% ± 16.9% | 82.4% ± 13.6% | 70.6% ± 16.3% | 50% ± 17.9% | 63.3% ± 17.2% | 86.7% ± 12.2% | 73.3% ± 15.8% |
| Holdout | 100% | 75% | 87.5% | 25% | 0% | 100% | 50% | 25% | 0% | 66.7% | 66.7% | 14% | 67% | 83.3% | 75% | 25% | |
| COBRE | Training | 100% | 57.7% ± 13.6% | 92.3% ± 7.3% | 96.2% ± 5.3% | 0% | 60% ± 13.4% | 88% ± 8.9% | 52% ± 13.7% | 0% | 57.7% ± 13.6% | 91.7% ± 7.6% | 92.9% ± 7.1% | 51% ± 13.7% | 54.9% ± 13.7% | 90.2% ± 8.2% | 74.5% ± 6.1% |
| Holdout | 100% | 60% | 100% | 100% | 0% | 70% | 100% | 60% | 0% | 63.6% | 100% | 100% | 50% | 65% | 100% | 80% | |
| Mind Research Network (MRN) | Training | 100% | N/A | 88.2% ± 11.5% | 100% | 0% | N/A | 84.6% ± 12.9% | 23.1% ± 15.1% | 0% | N/A | 84.6% ± 12.9% | 100% | 56.7% ± 17.7% | N/A | 86.7% ± 12.2% | 66.7% ± 16.9% |
| Holdout | 100% | N/A | 67% | 100% | 0% | N/A | 71% | 14% | 0% | N/A | 63% | 100% | 56% | N/A | 68.8% | 62.5% | |
| Stanford | Training | 100% | 73.3% ± 15.8% | 80% ± 14.3% | 100% | 0% | 66.7% ± 16.9% | 86.7% ± 12.2% | 33.3% ± 16.9% | 0% | 71.4% ± 16.2% | 81.3% ± 14% | 100% | 50% ± 17.9% | 70% ± 16.4% | 83.3% ± 13.3% | 66.7% ± 16.9% |
| Holdout | 100% | 80% | 80% | 100% | 0% | 40% | 60% | 0% | 0% | 67% | 75% | 0% | 50% | 60% | 70% | 50% | |
| University of Michigan (UM_2) | Training | 100% | 69.2% ± 20.2% | 100% | 100% | 0% | 42.9% ± 21.7% | 71.4% ± 19.8% | 14.3% ± 15.3% | 0% | 42.9% ± 21.7% | 100% | 100% | 65% ± 20.9% | 60% ± 21.5% | 90% ± 13.1% | 70% ± 20.1% |
| Holdout | 100% | 66.6% | 100% | 83.3% | 0% | 25% | 50% | 0% | 0% | 20% | 100% | 0% | 60% | 30% | 80% | 50% | |