Table 6.
Results obtained with augmented data (AD) using the second dataset, the ANN machine learning algorithm, different data augmentation methods and alternative test data.
Data Augmentation Method | Accuracy (%) | ||
---|---|---|---|
100 AD | 500 AD | 2000 AD | |
Proposed | 92.67 | 93.00 | 91.33 |
Signal stretching | 84.33 | 84.00 | 87.67 |
Gaussian noise and signal stretching | 83.33 | 83.67 | 86.67 |
SMOTE | 88.33 | 88.67 | 85.67 |
Gaussian noise | 84.00 | 87.33 | 85.67 |