Table 21.
Results for the combined sets of attributes.
Attributes | Model | Accuracy | Precision | Recall | F1 Score |
---|---|---|---|---|---|
Physical Appearance | SVC | 71.0% | 72.8% | 71.2% | 0.708 |
KNN | 75.4% | 76.8% | 75.3% | 0.750 | |
AdaBoost | 72.3% | 73.5% | 72.1% | 0.719 | |
Physical appearance + Space and Environment Attributes | SVC | 83.9% | 86.1% | 83.4% | 0.831 |
KNN | 81.8% | 86.4% | 81.2% | 0.807 | |
AdaBoost | 76.3% | 78.4% | 75.8% | 0.755 | |
Physical Appearance + Space and Environment Attributes + Body Gestures and Actions | SVC | 91.3% | 92.3% | 91.2% | 0.912 |
KNN | 94.7% | 95.5% | 94.4% | 0.946 | |
AdaBoost | 92.5% | 93.2% | 92.7% | 0.925 | |
All Attributes | SVC | 94.6% | 95.3% | 94.6% | 0.945 |
KNN | 95.6% | 96.2% | 95.7% | 0.956 | |
AdaBoost | 93.5% | 94.1% | 93.4% | 0.934 |