Table 3 Model robustness and noise resistance analysis results.
From: The analysis of motion recognition model for badminton player movements using machine learning
Noise level | Algorithm | ACC | P | R | F1 |
|---|---|---|---|---|---|
Low noise | SVM | 0.902 | 0.91 | 0.89 | 0.90 |
3D CNN | 0.915 | 0.92 | 0.91 | 0.91 | |
QCNN | 0.945 | 0.95 | 0.94 | 0.94 | |
Medium noise | SVM | 0.818 | 0.84 | 0.80 | 0.82 |
3D CNN | 0.862 | 0.88 | 0.85 | 0.86 | |
QCNN | 0.913 | 0.92 | 0.90 | 0.91 | |
High noise | SVM | 0.715 | 0.75 | 0.70 | 0.72 |
3D CNN | 0.78 | 0.80 | 0.76 | 0.78 | |
QCNN | 0.882 | 0.89 | 0.87 | 0.88 |
