From: ECG signal classification based on deep CNN and BiLSTM
Method | F1-score | Accurary | |||
---|---|---|---|---|---|
Normal | AF | Other | Overall | ||
Multi-SVM [22] | 0.90 | 0.81 | 0.72 | 0.81 | |
Double-layer independent CNN [23] | 0.91 | 0.83 | O.72 | 0.82 | |
21-layer 1D CNN [24] | 0.919 | 0.858 | 0.816 | 0.864 | |
XGBoost and LSTMs stacked by LDA [25] | 0.953 | 0.838 | 0.850 | 0.880 | |
Decision tree ensemble [26] | 0.889 | 0.791 | 0.702 | 0.80 | |
8CSL [29] | – | 0.895 | – | – | |
Article method | 0.931 | 0.864 | 0.830 | 0.891 | 0.893 |