Google Cloud Certified Professional Machine Learning Engineer Practice Mode

Community Discussion

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

• Chose C1

  Below explanation from AI Your task is a 3‑class single‑label image classification problem: drivers_license, passport, credit_card Each image belongs to exactly one of these classes.This means the model should output a probability distribution over 3 classes, e.g.:[0.85, 0.10, 0.05] For this setup, the correct loss is:✅ Categorical cross‑entropy Used when: You have multiple classes Each sample belongs to one class Labels are one‑hot encoded, e.g.drivers_license → [1,0,0] This matches your scenario perfectly.

• Chose D1

  model is set up for mutually exclusive single-label classification (meaning an image can only be one thing at a time)

• Chose C1

  Because sparse categorical cross-entropy is for integer values.