Skip to main content

Table 2 Common metrics in AI model evaluation

From: Application of machine learning in ophthalmic imaging modalities

Evaluation metricsDefinitions
AccuracyThe proportion of both positives and negatives that are correctly identified; the higher the accuracy, the better the classifier
Sensitivity/RecallThe proportion of positives that are correctly identified
SpecificityThe proportion of negatives that are correctly identified
PrecisionThe proportion of positives that are correctly identified among all positive identified samples
Kappa valueTo show the actual agreement between two sets of observations
Dice coefficient/F1 scoreHarmonic average of the precision and recall, where an F1 score reaches its best value at 1 and worst at 0
\