Table 6 Comparison of developed model with traditional machine learning classifiers and recent deep learning models.
From: Adaptive deep SVM for detecting early heart disease among cardiac patients
Deep learning technique | ||||
|---|---|---|---|---|
Methods | GCN21 | RawECGNet22 | O-SBGC-LSTM3 | EAVSRO-AD-SVM |
Accuracy | ||||
Adam | 89.35 | 91.15 | 88.6 | 96.37681 |
SGD | 87.6 | 90.425 | 87.975 | 96.58863 |
RMSprop | 88.875 | 89.975 | 87.425 | 96.20553 |
Adagrad | 89.675 | 91.45 | 90.625 | 96.95652 |
Gradient Descent | 88.725 | 91.475 | 90.525 | 96.70807 |
Precision | ||||
Adam | 89.3 | 91.3 | 88.8 | 88.75171 |
SGD | 87.3 | 90.3 | 87.9 | 89.87138 |
RMSprop | 89.4 | 89.9 | 87.4 | 89.88327 |
Adagrad | 89.8 | 91.6 | 90.7 | 91.48936 |
Gradient Descent | 89.1 | 91.7 | 90.5 | 91.38462 |
MCC | ||||
Adam | 0.741 | 0.782 | 0.725 | 98.4197 |
SGD | 0.702 | 0.765 | 0.711 | 98.35304 |
RMSprop | 0.732 | 0.755 | 0.699 | 97.81746 |
Adagrad | 0.749 | 0.789 | 0.770 | 98.36066 |
Gradient Descent | 0.729 | 0.790 | 0.768 | 98.05447 |
Machine learning technique | ||||
|---|---|---|---|---|
Optimizers | TML-EML13 | Machine learning- based framework26 | LS-SVM27 | SVM29 |
Accuracy | ||||
Adam | 87.600 | 91.275 | 89.400 | 96.37681 |
SGD | 89.425 | 89.775 | 89.150 | 96.58863 |
RMSprop | 88.150 | 88.650 | 88.575 | 96.20553 |
Adagrad | 89.425 | 91.200 | 90.975 | 96.95652 |
Gradient Descent | 88.525 | 91.850 | 90.475 | 96.70807 |
Precision | ||||
Adam | 87.400 | 91.200 | 89.200 | 88.75171 |
SGD | 89.800 | 89.600 | 89.100 | 89.87138 |
RMSprop | 88.100 | 88.300 | 88.700 | 89.88327 |
Adagrad | 89.500 | 91.200 | 91.000 | 91.48936 |
Gradient Descent | 88.600 | 92.100 | 90.400 | 91.38462 |
MCC | ||||
Adam | 0.702 | 0.785 | 0.742 | 98.4197 |
SGD | 0.744 | 0.750 | 0.737 | 98.35304 |
RMSprop | 0.715 | 0.725 | 0.725 | 97.81746 |
Adagrad | 0.743 | 0.783 | 0.778 | 98.36066 |
Gradient Descent | 0.723 | 0.799 | 0.766 | 98.05447 |
