Table 6.
The highest experiment outcomes of the binary and multi-class classifications for GWO-ELM approach using APTOS-2019 and IDRiD datasets.
| APTOS-2019 dataset | |||||||
|---|---|---|---|---|---|---|---|
| Number of class | Accuracy | Precision | Recall | Specificity | MCC | F-measure | G-mean |
| 5 | 96.21 | 90.53 | 90.53 | 97.63 | 88.16 | 90.53 | 90.53 |
| 2 | 99.47 | 99.34 | 100.00 | 97.44 | 98.38 | 99.67 | 99.67 |
| IDRiD dataset | |||||||
| Number of class | Accuracy | Precision | Recall | Specificity | MCC | F-measure | G-mean |
| 5 | 96.15 | 90.38 | 90.38 | 97.60 | 87.98 | 90.38 | 90.38 |
| 2 | 99.04 | 100.00 | 98.59 | 100.00 | 97.82 | 99.29 | 99.29 |