MLS-C01 Question #135: Real Exam Question with Answer & Explanation

Question

A financial company is trying to detect credit card fraud. The company observed that, on average, 2% of credit card transactions were fraudulent. A data scientist trained a classifier on a year's worth of credit card transactions data. The model needs to identify the fraudulent transactions (positives) from the regular ones (negatives). The company's goal is to accurately capture as many positives as possible. Which metrics should the data scientist use to optimize the model? (Choose two.)

Options

• ASpecificity
• BFalse positive rate
• CAccuracy
• DArea under the precision-recall curve
• ETrue positive rate

Explanation

{"question_number": 4, "correct_answer": "D, E", "explanation": "With only 2% of transactions being fraudulent, the dataset is highly imbalanced. The company's goal is to maximize detection of true fraud cases (positives). True Positive Rate (TPR), also known as recall or sensitivity (option E), directly measures the fraction of actual fraud cases correctly identified - exactly what the company wants to maximize. Area Under the Precision-Recall Curve (AUPRC, option D) is the ideal summary metric for imbalanced classification problems because it focuses on the minority positive class and captures the trade-off between precision and recall across all thresholds. Accuracy (C) is misleading on imbalanced data - a model that labels everything as non-fraud would achieve 98% accuracy while catching zero fraud. Specificity (A) and False Positive Rate (B) measure performance on the majority (non-fraud) class, which is not the primary concern here.", "generated_by": "claude-sonnet", "llm_judge_score": 3}

No community discussion yet for this question.