Table 4.
The comparison study for COVID-19 classification using deep and ML models.
| Study | Dataset | Model Used | Classification Accuracy |
|---|---|---|---|
| Narin et al. [43] | 2-class: 50 COVID-19/50 normal |
Transfer learning with ResNet50 and Inception-v3 | 98% |
| Panwar et al. [44] | 2-class: 142 COVID-19/ 142 normal |
nCOVnet CNN | 88% |
| Altan et al. [45] | 3-class: 219 COVID-19 1341 norma l1345 pneumonia viral |
2D curvelet transform, chaotic salp swarm algorithm (CSSA), EfficientNet-B0 | 99% |
| Chowdhury et al. [46] | 3-class: 423 COVID-19 1579 normal 1485 pneumonia viral |
Transfer learning with CheXNet | 97.7% |
| Wang and Wong [47] | 3-class: 358 COVID-19/5538 normal/8066 pneumonia |
COVID-Net | 93.3% |
| Kumar et al. [48] | 3-class: 62 COVID-19/1341 normal/1345 pneumonia |
ResNet1523 features and XGBoost classifier | 90% |
| Sethy and Behera [49] | 3-class: 127 COVID-19/127 normal/127 pneumonia |
ResNet50 features and SVM | 95.33% |
| Ozturk et al. [50] | 3-class: 125 COVID-19/500 normal 500 pneumonia |
DarkCovidNet CNN | 87.2% |
| Khan et al. [51] | 4-class: 284 COVID-19/310 normal/330 pneumonia bacterial/327 pneumonia viral |
CoroNet CNN | 89.6% |
| Tanvir Mahmud et al. [52] | 4-class: 305 COVID-19 + 305 Normal + 305 Viral Pneumonia + 305 Bacterial Pneumonia |
StackedMulti-resolutionCovXNet | 90.3% |
| Proposed CoVIRNet DL model | 4-class: 284 COVID-19/310 normal/330 pneumonia bacterial/327 pneumonia viral |
Multiscale features CoVIRNet | 95.78% |
| Proposed CoVIRNet DL model with RF | 4-class: 284 COVID-19/310 normal/330 pneumonia bacterial/327 pneumonia viral |
Multiscale features CoVIRNet+ RF | 97.29% |