Research Article
Diagnosis of Alzheimer’s Disease with Extreme Learning Machine on Whole-Brain Functional Connectivity
Table 1
10 times 10-fold cross-validation results for AD and CN classification.
| | No. of experiments | ELM classifiers | Optimal ELM classifiers | Parallel ELM classifiers | | Amount | Mean Accuracy ± STD | Worst accuracy (%) | Amount | Mean Accuracy ± STD | Worst accuracy (%) | Amount | Mean accuracy ± STD | Worst accuracy (%) |
| | 1 | 2362 | 95.50% ± 4.91% | 70.00 | 771 | 95.82% ± 4.88% | 70.00 | 10 | 96.00% ± 6.15% | 80.00 | | 2 | 3024 | 95.60% ± 4.75% | 75.00 | 1029 | 95.68% ± 4.62% | 75.00 | 10 | 96.50% ± 3.37% | 90.00 | | 3 | 2090 | 94.97% ± 4.40% | 75.00 | 662 | 95.38% ± 4.11% | 80.00 | 10 | 97.00% ± 3.50% | 90.00 | | 4 | 3248 | 94.01% ± 5.15% | 70.00 | 1052 | 94.14% ± 5.00% | 80.00 | 10 | 96.50% ± 4.74% | 85.00 | | 5 | 2434 | 96.86% ± 4.36% | 75.00 | 765 | 97.18% ± 3.96% | 75.00 | 10 | 98.50% ± 2.42% | 95.00 | | 6 | 1748 | 95.05% ± 4.67% | 80.00 | 545 | 95.41% ± 4.41% | 80.00 | 10 | 97.50% ± 3.54% | 90.00 | | 7 | 2680 | 95.17% ± 4.31% | 80.00 | 855 | 95.26% ± 4.30% | 80.00 | 10 | 97.50% ± 3.54% | 90.00 | | 8 | 2110 | 94.45% ± 4.74% | 75.00 | 702 | 94.60% ± 4.59% | 75.00 | 10 | 96.00% ± 4.59% | 90.00 | | 9 | 1580 | 95.08% ± 4.45% | 75.00 | 558 | 95.21% ± 4.17% | 80.00 | 10 | 96.50% ± 2.42% | 95.00 | | 10 | 2558 | 94.42% ± 5.00% | 75.00 | 846 | 94.77% ± 4.80% | 80.00 | 10 | 96.50% ± 4.74% | 90.00 | | Mean | 2383.4 | 95.11% ± 4.67% | 75.00 | 778.5 | 95.35% ± 4.49% | 77.50 | 10 | 96.85% ± 3.90% | 89.50 |
|
|
