MLS-C01 Question 217
Single answerA data scientist trains a binary classification model on imbalanced data where 95% of the instances belong to the negative class and 5% to the positive class. During evaluation, the model achieves a high accuracy of 96%, but stakeholders are concerned that the model may not perform well in identifying the positive class. Which metric should the data scientist prioritize to better evaluate the model's performance on the minority class?
- A
Precision
- B
Recall
- C
Accuracy
- D
F1 Score
Show answer and explanation
Correct answer: B
Explanation
In imbalanced datasets, metrics like accuracy can be misleading as they may give high values despite poor performance on the minority class. Since stakeholders are concerned about the model's ability to identify the positive class, recall is the most appropriate metric to prioritize. Recall directly measures the model's sensitivity to detecting positive instances, making it the most relevant metric in this scenario.
- A. Incorrect.
Precision measures the proportion of true positive predictions out of all positive predictions made by the model. While useful, it does not directly address the concern of the model's ability to identify positive instances (minority class) overall.
- B. Correct.
Recall measures the proportion of true positive instances that are correctly identified by the model. Since the stakeholders are concerned about the model's ability to identify the positive class (minority class), recall is the most relevant metric to evaluate here.
- C. Incorrect.
Accuracy measures the proportion of correctly classified instances out of all instances. However, in imbalanced datasets, accuracy can be misleading as it can be dominated by the majority class, leading to the high accuracy observed in this scenario, even if the minority class is poorly predicted.
- D. Incorrect.
F1 Score is the harmonic mean of precision and recall. While it provides a balanced view of both metrics, stakeholders specifically want to evaluate the model's ability to identify the positive class, which makes recall more directly relevant in this case.