Table 1. The binary classification prediction scores for Herlev and SiPaKMeD Cervical Cancer datasets under evaluation criteria, that is, Accuracy, Precision, Recall, F1-Score, and Kappa Score.
The table demonstrates the scores for various binary classifiers and CNN models, that is, K-Nearest Neighbour, Support Vector Machine, Stochastic Gradient Descent, Random Forest, ResNet-34 (Baseline), and EfficientNet-B3 (Baseline).
| Model | Accuracy (%) | Precision (%) | Recall (%) | F1-Score (%) | K-Score (%) |
|---|---|---|---|---|---|
| Herlev Dataset | |||||
| K-Nearest Neighbour | 78.142 | 79.268 | 95.588 | 86.666 | – |
| Support Vector Machine | 76.502 | 76.271 | 99.264 | 86.261 | – |
| Stochastic Gradient Descent | 71.584 | 78.378 | 85.294 | 81.690 | – |
| Random Forest | 74.863 | 75.568 | 97.794 | 85.256 | – |
| SIPaKMeD Dataset | |||||
| ResNet-34 | 98.919 | 99.291 | 98.925 | 98.918 | 99.103 |
| EfficientNet-B3 | 99.011 | 99.157 | 98.896 | 99.026 | 98.879 |