Databricks Machine Learning Associate Question 167
Single answerYou are building a binary classification model in Databricks for detecting fraudulent transactions. After evaluating the model, you find that the precision at the default threshold of 0.5 is too low, and you want to increase it. How can adjusting the designated threshold help improve the model's precision?
- A
Increase the threshold to classify fewer transactions as fraudulent.
- B
Decrease the threshold to classify more transactions as fraudulent.
- C
Keep the threshold at 0.5 and retrain the model to optimize precision.
- D
Adjust the threshold to balance precision and recall equally.
Show answer and explanation
Correct answer: A
Explanation
Adjusting the designated threshold is a common technique in classification tasks to modify model performance metrics like precision and recall. When you increase the threshold, the model becomes stricter in classifying positive samples, reducing false positives and increasing precision. However, this typically comes at the cost of recall, as fewer true positives are captured.
- A. Correct.
Increasing the threshold reduces the number of transactions classified as fraudulent, which decreases false positives and thus improves precision.
- B. Incorrect.
Decreasing the threshold increases the number of transactions classified as fraudulent, which may increase false positives and lower precision.
- C. Incorrect.
Keeping the threshold at 0.5 and retraining the model does not directly address the precision issue caused by the threshold setting.
- D. Incorrect.
Balancing precision and recall equally may not align with the goal of improving precision specifically.