Table 6 Comparative performance of DGOA-ensemble model against traditional optimization and classification methods.
From: An enhanced diabetic retinopathy detection approach using optimized deep learning technique
Feature selection | Classifier | Accuracy (%) | Precision | Recall | F1-score | AUC-ROC | Training time (s) | Inference time (s) |
|---|---|---|---|---|---|---|---|---|
GA | SVM | 85.2 | 0.84 | 0.82 | 0.83 | 0.88 | 120 | 1.5 |
RF | 86.5 | 0.85 | 0.84 | 0.84 | 0.89 | 150 | 1.8 | |
CNN | 89.1 | 0.88 | 0.87 | 0.87 | 0.92 | 200 | 2.5 | |
PSO | SVM | 83.8 | 0.82 | 0.81 | 0.81 | 0.86 | 130 | 1.6 |
RF | 85.0 | 0.84 | 0.83 | 0.83 | 0.88 | 160 | 1.9 | |
CNN | 87.9 | 0.87 | 0.86 | 0.86 | 0.91 | 210 | 2.7 | |
GOA | SVM | 86.0 | 0.85 | 0.83 | 0.84 | 0.89 | 125 | 1.4 |
RF | 88.2 | 0.87 | 0.86 | 0.86 | 0.91 | 170 | 2.0 | |
CNN | 90.3 | 0.89 | 0.88 | 0.89 | 0.93 | 220 | 2.8 | |
DGOA (Proposed) | Ensemble | 94.1 | 0.93 | 0.92 | 0.93 | 0.97 | 190 | 1.9 |
