MLS-C01 Question 233
Select 3You are building a binary classification model to predict whether a customer will churn (1) or not churn (0). After evaluating the model on a validation dataset, you obtain the following confusion matrix:
| Predicted: 0 | Predicted: 1 | |
|---|---|---|
| Actual: 0 | 800 | 200 |
| Actual: 1 | 50 | 950 |
Which of the following statements are correct based on this confusion matrix?
- A
The precision of the model is 0.826.
- B
The recall of the model is 0.95.
- C
The accuracy of the model is 0.88.
- D
False positives are higher than false negatives.
- E
The model is better at predicting 'not churn' compared to 'churn'.
Show answer and explanation
Correct answers: A, B, C
Explanation
To interpret a confusion matrix, it is essential to understand the definitions of precision, recall, accuracy, false positives, and false negatives. Here, the precision (0.826), recall (0.95), and accuracy (0.88) are calculated directly from the confusion matrix. False positives (200) are fewer than false negatives (50), and the model performs better at predicting 'churn' rather than 'not churn'.
- A. Correct.
Correct. Precision is calculated as True Positives / (True Positives + False Positives). Here, precision = 950 / (950 + 200) = 950 / 1150 ≈ 0.826.
- B. Correct.
Correct. Recall is calculated as True Positives / (True Positives + False Negatives). Here, recall = 950 / (950 + 50) = 950 / 1000 = 0.95.
- C. Correct.
Correct. Accuracy is calculated as (True Positives + True Negatives) / Total samples. Accuracy = (950 + 800) / (950 + 800 + 200 + 50) = 1750 / 2000 = 0.88.
- D. Incorrect.
Incorrect. False positives (200) are less than false negatives (50).
- E. Incorrect.
Incorrect. The model exhibits better performance in predicting 'churn' (class 1) based on the precision (0.826) and recall (0.95) for this class.