A fraud detection system flags suspicious credit card transactions for employee review. The company wants to reduce the time spent reviewing flagged cases that are actually legitimate. Which evaluation metric best supports this goal?

Precision is the most suitable metric because it measures the proportion of correctly identified positive cases (true positives) out of all cases predicted as positive (true positives + false positives). In this scenario, a false positive means a legitimate transaction was flagged as suspicious, leading to unnecessary review. Maximizing precision directly reduces the number of false positives, thereby minimizing the time employees spend reviewing legitimate transactions. Recall measures the proportion of true positives out of all actual positive cases (true positives + false negatives). While important for not missing actual fraud, optimizing recall alone could increase false positives. Accuracy measures overall correct predictions but doesn't specifically address the cost of false positives. A lift chart is a visualization tool, not a single evaluation metric.