PMLE Question 69

Community Discussion

Written by other learners. Comments can disagree with the marked answer — check the official documentation when in doubt.

• Chose A1

  In AI safety tasks, the "bad" thing is usually labeled as the Positive class. Therefore, a False Negative is when a bad photo is predicted as Negative (safe) and allowed on the site. The prompt says: "ensure that the application does not falsely accept a non-compliant picture." This is the literal definition of minimizing False Negatives. The mathematical way to minimize False Negatives is to tell the model to maximize Recall. Vertex AI AutoML allows you to upload your images and simply select "Optimize for Recall" from a dropdown menu, requiring zero custom coding.

• Chose D1

  ❌ A. Optimize Recall ----> Incorrect Metric: Optimizing recall for the positive class (compliant) minimizes False Negatives (rejecting a good photo). ❌ B. Optimize F1 Score---->Too Balanced: F1 is the harmonic mean of Precision and Recall. It seeks a balance between both types of errors. It does not provide the specific "strictness" required to prioritize the prevention of non-compliant photo acceptance. ✅D. three times as many examples of pictures that do not meet the profile photo requirements.:----> By oversampling non-compliant photos, you teach the model to better distinguish and not mistakenly accept non-compliant photos, thereby reducing false positives.

• Chose D1

  A is, definitely, not the answer, since we need to minimize false positive. B is used for balanced dataset which is not the case for this problem. C doesn't make sense for this problem D is the answer since we need oversampling of minority class (in this case, Picture non-compliant) to train the model for better prediction of minority class