A company wants to detect credit card fraud. The company has observed that an average of 2% of credit card transactions are fraudulent. A data scientist trains a classifier on a year's worth of credit

Specificity measures the proportion of actual negative cases (non-fraudulent transactions) that are correctly identified, which is not the primary objective of capturing fraudulent transactions.

False positive rate measures the proportion of non-fraudulent transactions that are incorrectly classified as fraudulent, and while minimizing it is important, it does not directly address the goal of capturing as many fraudulent transactions as possible.

Accuracy can be misleading in imbalanced datasets because a model that always predicts the majority class will still show high accuracy, failing to correctly identify the critical minority class (fraud).

The F1 score is a harmonic mean of precision and recall, which is particularly useful for imbalanced datasets like credit card fraud where both minimizing false positives and false negatives are important, balancing the need to identify fraud without excessive false alarms.

True Positive Rate (Recall) measures the proportion of actual fraudulent transactions that are correctly identified by the model, directly addressing the requirement to "accurately capture as many fraudulent transactions as possible."