Table 1 Performance of each model (mean, \(95\%\) confidence interval)
Method | AUC | F1 Score | Accuracy | ECE |
|---|---|---|---|---|
(larger is better) | (larger is better) | (larger is better) | (smaller is better) | |
Random | 0.49, [0.45 to 0.53] | 0.39, [0.35 to 0.43] | 0.48, [0.44 to 0.52] | 0.11, [0.10 to 0.12] |
No Pre-Train | 0.56, [0.50 to 0.62] | 0.32, [0.23 to 0.41] | 0.69, [0.67 to 0.71] | 0.12, [0.10 to 0.14] |
ImageNet | 0.67, [0.71 to 0.63] | 0.53, [0.49 to 0.57] | 0.74, [0.71 to 0.77] | 0.12, [0.09 to 0.15] |
C2L | 0.73, [0.70 to 0.76] | 0.62, [0.57 to 0.67] | 0.79, [0.77 to 0.81] | 0.15, [0.13 to 0.17] |
Proposed | 0.76, [0.71 to 0.81] | 0.64, [0.58 to 0.70] | 0.78, [0.75 to 0.81] | 0.09, [0.07 to 0.11] |