Databricks Machine Learning Associate Question 607
Single answerYou are building a machine learning model to predict whether a customer will churn (binary classification). The business objective is to minimize the number of customers incorrectly classified as non-churners (false negatives), as this would lead to a loss in revenue. Which evaluation metric is the most appropriate to optimize for this objective?
- A
Accuracy
- B
Precision
- C
Recall
- D
F1-score
Show answer and explanation
Correct answer: C
Explanation
In this scenario, the primary business objective is to minimize the number of false negatives (customers incorrectly classified as non-churners). Recall is the most appropriate metric because it directly measures the ability of the model to correctly identify all actual churners, thereby reducing the number of false negatives. While other metrics like accuracy, precision, and F1-score have their own uses, they do not specifically align with the business goal in this case.
- A. Incorrect.
Accuracy measures the overall correctness of the model. While it is useful in balanced datasets, it is not appropriate for imbalanced scenarios, especially when false negatives have a high cost, as it does not specifically account for this issue.
- B. Incorrect.
Precision measures the proportion of true positives out of all predicted positives. It focuses on reducing false positives and is not suitable when minimizing false negatives is the primary objective.
- C. Correct.
Recall measures the proportion of true positives out of all actual positives. It directly focuses on reducing false negatives, making it the most appropriate metric for this scenario where minimizing false negatives is critical.
- D. Incorrect.
F1-score is the harmonic mean of precision and recall. While it balances both metrics, it is not ideal when the business concern is heavily skewed towards minimizing false negatives, as recall alone is more focused on this objective.