Table 11 Performance of the model after reducing the learning rates for backbone and projection head.

From: A hybrid learning framework for automated multiclass electrocardiogram classification with SimCardioNet

Class

Precision

Recall

F1-score

ECG MI Patients

0.92

0.90

0.91

History of MI

0.90

0.89

0.90

Abnormal Heartbeat

0.93

0.91

0.92

Normal ECG

0.91

0.92

0.91

Overall accuracy

–

0.91

  1. Significant values are in bold.
Back to article page