Table 3. 5-fold CV results for classification accuracy (and standard deviation) using various numbers of independent components.
# ICs (SVM GAMMA) | 5-Fold CV Accuracy | 5-Fold CV STD | |
---|---|---|---|
ADHD | 30 (6.4) | 0.6216 | 0.0399 |
35 (3.2) | 0.6378 | 0.0374 | |
40 (1.6) | 0.6432 | 0.0390 | |
45 (3.2) | 0.6450 | 0.0291 | |
50 (3.2) | 0.6360 | 0.0261 | |
55 (3.2) | 0.6378 | 0.0075 | |
Autism | 30(51.2) | 0.5925 | 0.0158 |
35 (51.2) | 0.5825 | 0.0049 | |
40 (51.2) | 0.6159 | 0.0078 | |
45 (51.2) | 0.6225 | 0.0210 | |
50 (51.2) | 0.6000 | 0.0130 | |
55 (51.2) | 0.5987 | 0.0059 |
Note: SVM GAMMA refers to the RBF kernel variance parameter .