Table 3. Confusion matrix for binary classification.
During the calculation of the metrics, the actual and estimated true/false numbers are determined using the confusion matrix.
| Predicted:Mild | Predicted:Moderate | Predicted:Severe | Predicted:Healthy | |
|---|---|---|---|---|
| Actual:Mild | TP | FN | FN | FN |
| Actual:Moderate | FP | TP | FN | FN |
| Actual:Severe | FP | FP | TP | FN |
| Actual:Healthy | FP | FP | FP | TP |