Table 7 the performance of various damaged models during ablation study on classifying COVID-19 and normal samples.
From: Potential diagnostic application of a novel deep learning- based approach for COVID-19
Model | Accuracy on detecting COVID-19 cases (%) | Accuracy on detecting normal cases (%) | Total accuracy (%) |
|---|---|---|---|
Main model | 100 | 100 | 100 |
Damaged model I | 15.38 | 100 | 47.62 |
Damaged model II | 100 | 62.5 | 85.71 |
Damaged model III | 100 | 0 | 61.90 |
Damaged model IV | 100 | 87.5 | 95.24 |
Damaged model V | 100 | 100 | 100 |
