Table 3.
Classification performance among single coordinate position and the xyz position. This accuracy is obtained for a window size of 125.
| Classifier/Acc. | Cp(x) | Cp(y) | Cp(z) | Cp (x, y, z) |
|---|---|---|---|---|
| Bagged Tree | 100 | 100 | 100 | 100 |
| Boosted Tree | 99.8 | 33.3 | 99.5 | 100 |
| Coarse Gaussian Support Vector Machine (SVM) | 73.1 | 78.3 | 63.5 | 52.3 |
| Coarse K-Nearest Neighbor (KNN) | 90.4 | 96.8 | 90.4 | 86.5 |
| Cosine KNN | 99.6 | 99.9 | 100 | 99.5 |
| Cubic SVM | 99.8 | 93 | 86.8 | 92.5 |
| Complex Tree | 99.7 | 99.9 | 99.4 | 99.8 |
| Cubic KNN | 99.6 | 99.8 | 96.6 | 99.3 |
| Fine Gaussian SVM | 99.8 | 93 | 82.9 | 92.6 |
| Fine KNN | 100 | 100 | 99.4 | 100 |
| Linear Discriminant | 62.1 | 61.1 | 60 | 53.5 |
| Linear SVM | 67.3 | 77.8 | 57.3 | 52.9 |
| Medium Gaussian SVM | 97.1 | 93 | 77 | 87.6 |
| Medium KNN | 99.6 | 99.8 | 97.2 | 99.4 |
| Medium Tree | 99.7 | 99.9 | 93.7 | 96.3 |
| Quadratic Discriminant | 49.5 | 61.5 | 46.2 | 43.6 |
| Quadratic SVM | 99 | 98.2 | 81.3 | 85 |
| Random Under Sampling (RUS) Boosted Tree | 99.7 | 33.3 | 93.7 | 96.3 |
| Subspace KNN | 99.6 | 99.4 | 94 | 99.3 |
| Subspace Discriminant | 67.5 | 63.6 | 62.8 | 51.5 |
| Simple Tree | 88.3 | 97.2 | 77.8 | 74.5 |
| Weighted KNN | 100 | 99.9 | 98.7 | 99.9 |
| Mean Accuracy | 90.50 | 85.39 | 84.46 | 84.65 |