Google Professional Machine Learning Engineer
Get started today
Ultimate access to all questions.
Comments
Loading comments...
Ultimate access to all questions.
Your data science team is tasked with developing a machine learning model to predict customer churn. The team needs to experiment with various features, model architectures, and hyperparameters efficiently. They require a solution that not only automates the tracking of accuracy metrics across experiments but also allows for querying these metrics over time via an API to analyze trends and make informed decisions. The solution should minimize manual effort, support scalability, and ensure compliance with data governance policies. Which of the following solutions best meets these requirements? Choose the best option.
A
Utilize AI Platform Training for conducting experiments. Log all accuracy metrics directly to BigQuery, and use the BigQuery API for querying the metrics. This approach leverages Google Cloud's managed services for scalability and compliance.
B
Conduct experiments using AI Platform Training and record accuracy metrics in Cloud Monitoring. Access these metrics through the Monitoring API. This method provides real-time monitoring capabilities and integrates with Google Cloud's operations suite.
C
Perform experiments within AI Platform Notebooks and aggregate the results in a shared Google Sheets document. Query the results using the Google Sheets API. This option offers simplicity and ease of use for teams familiar with Google Workspace.
D
Implement Kubeflow Pipelines for running experiments. Automatically export metrics files and query the results through the Kubeflow Pipelines API. This solution provides comprehensive automation, scalability, and flexibility for experiment tracking and analysis.
E
Combine the use of AI Platform Training for experiment execution with Cloud Logging for metrics tracking. Query the logged metrics via the Logging API. This approach ensures detailed logging and querying capabilities within Google Cloud's ecosystem.