Table 2. Performance classification of a one-level system.
| Estimation of the best classifier | Test | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1R | DT | 1-NN | 5-NN | NB | RBF | SVM | LR | AdaB | Bag | RF | SVM | ||
| K3b | Accuracy (%) | 39.1 | 64.8 | 72.2 | 75.6 | 75.7 | 77.0 | 81.6* | 67.7 | 75.8 | 71.9 | 73.0 | 77.5±3.7 |
| FPR (%) | 75.0 | 52.0 | 48.3 | 48.7 | 35.7 | 33.0 | 31.6 | 42.3 | 37.3 | 44.0 | 47.0 | 29.2±4.9 | |
| L1b | Accuracy (%) | 45.8 | 62.6 | 62.8 | 70.3 | 70.3 | 70.5 | 73.5* | 60.3 | 69.3 | 68.7 | 68.4 | 74.4±3.2 |
| FPR (%) | 79.0 | 50.5 | 48.5 | 41.5 | 45.0 | 37.0 | 34.0 | 42.5 | 42.0 | 41.5 | 45.5 | 27.5±7.2 | |
| K6b | Accuracy (%) | 35.5 | 47.7 | 49.9 | 52.8 | 51.3 | 57.1 | 57.3* | 48.5 | 56.2 | 56.1 | 55.4 | 55.6±3.9 |
| FPR (%) | 55.0 | 55.0 | 51.0 | 55.0 | 45.0 | 41.0 | 44.5 | 53.0 | 41.5 | 38.5 | 39.0 | 36.2±10.0 | |
| Aver. | Accuracy (%) | 40.0 | 59.3 | 63.1 | 67.6 | 67.2 | 69.5 | 72.3* | 60.1 | 68.3 | 66.5 | 66.6 | 70.4±3.6 |
| FPR (%) | 70.4 | 52.4 | 49.1 | 48.4 | 41.0 | 36.4 | 36.0 | 45.4 | 39.9 | 41.7 | 44.3 | 30.7±7.0 | |
* shows the highest accuracy for each subject when estimating the best classifier using 5-run, 10-CV. The standard deviation is also given for the test results.