Table 2 Formulas for the statistical features extracted from the confusion matrix, where \(n_{TP}\), \(n_{FP}\), \(n_{TN}\) and \(n_{FN}\) represent the number of true positives, false positives, true negatives and false negatives, respectively.
From: Deep learning for the quality control of thermoforming food packages
Measure | Formula |
|---|---|
Accuracy (A) | \(\dfrac{n_{TP} + n_{TN}}{n_{TP} + n_{TN} + n_{FP} + n_{FN}}\) |
Precision (P) | \(\dfrac{n_{TP}}{n_{TP} + n_{FP}}\) |
Recall (R) | \(\dfrac{n_{TP}}{n_{TP} + n_{FN}}\) |
F-score (F) | \(2 \times \dfrac{Precision \times Recall}{Precision + Recall}\) |
False omission rate (FOR) | \(\dfrac{n_{FN}}{n_{FN} + n_{TN}}\) |
False discovery rate (FDR) | \(\dfrac{n_{FP}}{n_{FP} + n_{TP}}\) |
