Google Professional Machine Learning Engineer Question 124
Select 3Google Cloud PlatformYou are building a machine learning model on Google Cloud that processes healthcare data to predict patient outcomes. The dataset includes sensitive information such as patient names, medical records, and addresses. Which actions should you take to ensure compliance with privacy regulations and protect sensitive data while designing the system?
- A
Use Data Loss Prevention (DLP) API to identify and mask sensitive data in the dataset before processing.
- B
Store the entire dataset, including sensitive information, in a public Cloud Storage bucket for easy access during development.
- C
Implement encryption at rest and in transit for all sensitive data.
- D
Ensure that the machine learning model does not memorize or reproduce sensitive data during predictions.
- E
Share unmasked data with third-party collaborators to improve model performance.
Show answer and explanation
Correct answers: A, C, D
Explanation
When handling sensitive data, such as healthcare information, it is critical to adhere to privacy regulations and implement Google Cloud's best practices for security and compliance. This includes using tools like the Data Loss Prevention API to identify and mask sensitive data, encrypting data at rest and in transit, and ensuring that the model does not leak sensitive data. Avoid practices such as storing sensitive data in public buckets or sharing unmasked data with third parties, as these actions can lead to privacy violations and regulatory non-compliance.
- A. Correct.
Using the Data Loss Prevention (DLP) API is a best practice for identifying, masking, or redacting sensitive data to ensure compliance with privacy regulations such as HIPAA or GDPR.
- B. Incorrect.
Storing sensitive data in a public Cloud Storage bucket is a violation of privacy and security best practices. Sensitive data should never be stored in publicly accessible locations.
- C. Correct.
Encrypting sensitive data at rest and in transit ensures that unauthorized users cannot access the data, even if a breach occurs. This is a critical step in protecting sensitive information.
- D. Correct.
Ensuring that the model does not memorize or reproduce sensitive data during predictions is essential to prevent data leakage and maintain privacy compliance.
- E. Incorrect.
Sharing unmasked data with third parties is a serious privacy violation and could result in non-compliance with regulations such as GDPR or HIPAA. Proper anonymization or masking of sensitive data is required before sharing.