Table 8 Classification reports of DL models on PTBDB dataset.
From: An explainable deep learning framework for trustworthy arrhythmia detection from ECG signals
Data Balancing | Model | Class | Precision | Recall | F1-Score |
|---|---|---|---|---|---|
ADASYN | CNN | Normal | 98.44 | 99.58 | 99.01 |
Arrhythmia | 99.57 | 98.38 | 98.97 | ||
DNN | Normal | 94.62 | 98.00 | 96.28 | |
Arrhythmia | 97.88 | 94.29 | 96.05 | ||
SMOTE | CNN | Normal | 98.77 | 99.48 | 99.12 |
Arrhythmia | 99.47 | 98.76 | 99.12 | ||
DNN | Normal | 96.50 | 96.96 | 96.73 | |
Arrhythmia | 96.94 | 96.48 | 96.71 | ||
SMOTETomek | CNN | Normal | 98.96 | 99.24 | 99.10 |
Arrhythmia | 99.24 | 98.95 | 99.09 | ||
DNN | Normal | 96.72 | 96.90 | 96.81 | |
Arrhythmia | 96.90 | 96.71 | 96.81 | ||
ROS | CNN | Normal | 99.29 | 99.57 | 99.43 |
Arrhythmia | 99.57 | 99.29 | 99.43 | ||
DNN | Normal | 95.71 | 96.67 | 96.19 | |
Arrhythmia | 96.63 | 95.67 | 96.15 |
