Table 6.
Performance of cross-conditional classification of local features in sensor and source space using the framework explained in “Temporal evolution of internal mental states” section
| Classifier | Feature | Classification performance (%) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Necker cube (NC) | Stroboscopic alternative motion (SAM) | Structure from motion (SFM) | ||||||||
| Accuracy | Sensitivity | Specificity | Accuracy | Sensitivity | Specificity | Accuracy | Sensitivity | Specificity | ||
| SVM | Sensor | Training | 75.32 | 75.43 | 75.30 | 76.24 | 83.77 | 74.67 | ||
| Source | Training | 70.44 | 76.10 | 69.64 | 72.24 | 85.71 | 69.42 | |||
| Sensor | 67.94 | 66.03 | 68.21 | Training | 68.87 | 74.10 | 68.45 | |||
| Source | 67.87 | 63.51 | 68.49 | Training | 69.22 | 75.54 | 68.71 | |||
| Sensor | 72.44 | 72.05 | 72.50 | 72.69 | 73.03 | 72.67 | Training | |||
| Source | 70.78 | 63.64 | 71.80 | 71.69 | 71.96 | 71.68 | Training | |||
| ANN | Sensor | Training | 75.02 | 73.56 | 75.22 | 75.57 | 80.74 | 74.49 | ||
| Source | Training | 70.84 | 72.50 | 70.61 | 71.70 | 79.08 | 70.16 | |||
| Sensor | 70.07 | 71.65 | 69.85 | Training | 70.67 | 79.15 | 69.98 | |||
| Source | 62.28 | 65.06 | 61.88 | Training | 63.03 | 73.67 | 62.17 | |||
| Sensor | 72.95 | 71.52 | 73.15 | 73.13 | 71.70 | 73.19 | Training | |||
| Source | 64.80 | 66.48 | 64.56 | 65.03 | 74.10 | 64.62 | Training | |||