Recall vs. Sensitivity: Understanding Key Metrics in Machine Learning

It can be seen that recall and sensitivity are the same. FN refers to false negatives, which are classified as negative samples but are actually positive samples. FP refers to false positives, which are classified as positive samples but are actually negative samples. TN refers to true negatives, which are classified as negative samples and are actually negative samples. TP refers to true positives, which are classified as positive samples and are actually positive samples. Accuracy (overall accuracy) refers to the overall accuracy of the classification model (including the overall accuracy of all classes). Accuracy represents the proportion of correctly predicted samples among all predicted samples, and its formula is (8). Precision (single-class accuracy) refers to the accuracy of predicting 0. Precision represents the proportion of actual positive samples among the predicted positive samples. It can be seen that precision considers the proportion of positive samples predicted correctly. Its calculation formula can be obtained accordingly. The numerator of the calculation formula for accuracy and recall is TP, which refers to the number of positive samples predicted as positive samples. Therefore, they represent the precision and recall of positive samples. Accuracy mainly reflects the proportion of overall correct predictions.