Recall (Sensitivity or True Positive Rate)

Recall = TP / (TP + FN)

Recall, also known as sensitivity or true positive rate, is an evaluation metric used in classification tasks to measure the model’s ability to correctly identify positive instances. It quantifies the proportion of true positive predictions out of all actual positive instances in the dataset. For example, let’s consider a binary classification problem where the model correctly identifies 80 positive instances out of a total of 100 actual positive instances. However, it fails to identify the remaining 20 positive instances (false negatives). In this case, the recall would be 80 / (80 + 20) = 0.8 or 80%.

A recall is particularly important in scenarios where the consequences of false negatives are high. False negatives occur when the model incorrectly predicts a positive instance as negative, which can result in missing critical instances or opportunities. A high recall is desired in certain domains, such as medical diagnosis or anomaly detection. In medical diagnosis, missing the identification of a disease can lead to delayed treatment or serious consequences. In anomaly detection, failing to identify abnormal events or patterns can lead to security breaches or operational failures.

Classification evaluation metrics are quantitative measures used to assess the performance and accuracy of a classification model. These metrics provide insights into how well the model can classify instances into predefined classes or categories.

The commonly used classification evaluation metrics are: