Databricks Machine Learning Associate Question 608
Single answerYou are developing a machine learning model to predict whether a customer will churn or not for a subscription-based service. The business team has emphasized minimizing the number of customers incorrectly identified as non-churners (false negatives) because retaining at-risk customers is critical. Which evaluation metric is most appropriate for this objective?
- A
Accuracy
- B
Precision
- C
Recall
- D
F1 Score
Show answer and explanation
Correct answer: C
Explanation
The business objective is to minimize false negatives, as failing to identify churners could lead to lost revenue. Recall is the most appropriate metric because it directly focuses on the proportion of actual churners that are correctly identified, ensuring that at-risk customers are not overlooked.
- A. Incorrect.
Accuracy measures the proportion of correctly predicted instances out of all instances. While it's a useful metric in balanced datasets, it doesn't specifically address the need to minimize false negatives in this scenario.
- B. Incorrect.
Precision measures the proportion of true positives out of all predicted positives. While it helps reduce false positives, it doesn't directly prioritize minimizing false negatives, which is the key concern in this scenario.
- C. Correct.
Recall measures the proportion of true positives out of all actual positives. A higher recall ensures that most actual churners are correctly identified, minimizing false negatives, which aligns with the business objective.
- D. Incorrect.
F1 Score is the harmonic mean of precision and recall, which balances both metrics. However, it does not specifically prioritize recall over precision, making it less optimal for this scenario.