AWS Certified Machine Learning Engineer – Associate (MLA-C01) — Question 205

A company uses ML models to predict whether transactions are fraudulent. The company needs to identify as many fraudulent transactions as possible.

Which evaluation metric should the company use to evaluate the models to meet this requirement?

Answer options

• A. F1 score
• B. Area Under the ROC Curve (AUC)
• C. Precision
• D. Recall

Correct answer: D

Explanation

The correct answer is D, Recall, because it measures the ability of the model to identify all relevant cases, which in this context are the fraudulent transactions. While Precision (C) focuses on the accuracy of identified positives, it does not account for missed fraudulent transactions. The F1 score (A) is a balance between Precision and Recall, and AUC (B) assesses overall model performance without specifically prioritizing the identification of all fraudulent cases.