Table 4 Results obtained by the proposed framework on the three publicly available datasets used in this study, considering fivefold cross-validation scheme.
From: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology
Dataset | Fold | Accuracy (%) | Precision (%) | Recall (%) | F1-Score (%) |
|---|---|---|---|---|---|
SIPaKMeD 2-Class | 1 | 98.09 | 98.13 | 98.03 | 98.08 |
2 | 98.67 | 98.75 | 98.54 | 98.63 | |
3 | 97.80 | 97.85 | 97.73 | 97.79 | |
4 | 98.19 | 98.11 | 98.29 | 98.20 | |
5 | 100.00 | 100.00 | 100.00 | 100.00 | |
\({Avg} \pm {Std. Dev.}\) | \({{98.55}} \pm {{0.78}}\) | \({{98.57}} \pm {{0.77}}\) | \({{98.52}} \pm {{0.79}}\) | \({{98.54}} \pm {{0.78}}\) | |
SIPaKMeD 5-Class | 1 | 95.60 | 95.60 | 95.73 | 95.66 |
2 | 94.84 | 94.56 | 94.60 | 94.58 | |
3 | 95.34 | 95.21 | 95.34 | 95.27 | |
4 | 95.41 | 95.34 | 95.41 | 95.37 | |
5 | 95.96 | 96.00 | 95.81 | 95.90 | |
\({{Avg}} \pm {{Std. Dev.}}\) | \({{95.43}} \pm {{0.36}}\) | \({{95.34}} \pm {{0.48}}\) | \({{95.38}} \pm {{0.43}}\) | \({{95.36}} \pm {{0.45}}\) | |
Mendeley LBC | 1 | 98.96 | 98.96 | 98.96 | 98.96 |
2 | 99.48 | 99.12 | 99.48 | 99.30 | |
3 | 99.12 | 98.96 | 99.12 | 99.04 | |
4 | 99.12 | 99.12 | 99.12 | 99.12 | |
5 | 99.48 | 99.48 | 99.48 | 99.48 | |
\({{Avg}} \pm {{Std. Dev.}}\) | \({{99.23}} \pm {{0.21}}\) | \({{99.13}} \pm {{0.19}}\) | \({{99.23}} \pm {{0.21}}\) | \({{99.18}} \pm {{0.19}}\) |
