Table 8 Comparison between standard GOA and proposed DGOA for feature selection in DR detection.
From: An enhanced diabetic retinopathy detection approach using optimized deep learning technique
Method | Control strategy | Avg. selected features | Accuracy (%) | Sensitivity (%) | Specificity (%) | Convergence speed | Stability (Std. Dev.) |
|---|---|---|---|---|---|---|---|
GOA | Static parameters | 180 ± 12 | 90.3 | 88.5 | 91.5 | Slow | High variance |
DGOA (Proposed) | Dynamic multi-parameter adaptation | 94 ± 8 | 94.1 | 92.3 | 95.0 | Faster | Lower variance |
