Table 3.
Classification of color-coded images using convolutional neural networks
| Collection | CNN | Metric |
|---|---|---|
| 1 | 0.74 ± 0.01 | Accuracy |
| 0.74 ± 0.01 | F1 | |
| 0.61 ± 0.01 | MCC | |
| 2 | 0.72 ± 0.02 | Accuracy |
| 0.72 ± 0.02 | F1 | |
| 0.58 ± 0.03 | MCC | |
| 3 | 0.71 ± 0.02 | Accuracy |
| 0.71 ± 0.03 | F1 | |
| 0.56 ± 0.04 | MCC | |
| 4 | 0.73 ± 0.04 | Accuracy |
| 0.73 ± 0.04 | F1 | |
| 0.60 ± 0.06 | MCC | |
| 5 | 0.70 ± 0.04 | Accuracy |
| 0.70 ± 0.04 | F1 | |
| 0.55 ± 0.06 | MCC | |
| 6 | 0.72 ± 0.03 | Accuracy |
| 0.72 ± 0.03 | F1 | |
| 0.58 ± 0.04 | MCC | |
| 7 | 0.75 ± 0.03 | Accuracy |
| 0.75 ± 0.03 | F1 | |
| 0.62 ± 0.04 | MCC |
The table reports classification results for CNN models trained and tested on color-coded images. All values reported are averages ± standard deviations over 10 independent trials