Table 5.
Statistical study of PD detection with handwriting images using ILN-GNet over traditional methods for dataset 2.
| Methods | Mean | Minimum | Median | Std-Dev | Maximum |
|---|---|---|---|---|---|
| XGBoost | 0.760 | 0.732 | 0.759 | 0.022 | 0.790 |
| LSTM-CNN | 0.861 | 0.831 | 0.859 | 0.024 | 0.895 |
| LinkNet | 0.911 | 0.881 | 0.909 | 0.024 | 0.944 |
| GhostNet | 0.821 | 0.790 | 0.819 | 0.024 | 0.855 |
| GoogleNet | 0.861 | 0.831 | 0.859 | 0.024 | 0.895 |
| AlexNet | 0.880 | 0.851 | 0.878 | 0.023 | 0.911 |
| EfficientNet | 0.760 | 0.732 | 0.759 | 0.022 | 0.790 |
| CNN | 0.831 | 0.800 | 0.830 | 0.024 | 0.863 |
| Ensemble | 0.800 | 0.770 | 0.799 | 0.023 | 0.831 |
| ILN-GNet | 0.942 | 0.911 | 0.940 | 0.024 | 0.976 |