Answer-first summary for fast verification
Answer: Implement Kubeflow Pipelines to manage and run experiments, automatically track metrics, and use the Kubeflow Pipelines API for querying results., Combine the use of AI Platform Training for experiment execution and Kubeflow Pipelines for workflow management, storing metrics in both Cloud Monitoring and BigQuery for comprehensive analysis.
Kubeflow Pipelines (Option D) is the optimal choice for managing end-to-end machine learning workflows, enabling rapid experiment execution and automatic metric tracking. It significantly reduces manual effort by automating the tracking and reporting processes. Option E suggests a hybrid approach that leverages both AI Platform Training and Kubeflow Pipelines, offering flexibility and comprehensive metric analysis by storing data in both Cloud Monitoring and BigQuery. This approach is beneficial for teams requiring detailed analysis and scalability but involves more complexity and potentially higher costs. The correct answers are D and E, with D being the primary recommendation for its efficiency and automation capabilities.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
A data science team is working on a project that requires testing various features, model structures, and hyperparameters efficiently. The team needs a solution that not only allows for rapid experimentation but also ensures accurate tracking and reporting of experiment metrics with minimal manual effort. The solution should support scalability, cost-effectiveness, and compliance with data governance policies. Given these requirements, which of the following approaches is the BEST for the team to adopt? (Choose two correct options)
A
Utilize AI Platform Notebooks for running experiments and manually record the results in a shared Google Sheets file, then use the Google Sheets API for querying the results.
B
Deploy AI Platform Training for executing experiments, record accuracy metrics in Cloud Monitoring, and use the Monitoring API to query the results.
C
Use AI Platform Training to execute experiments, store accuracy metrics in BigQuery, and query the results via the BigQuery API.
D
Implement Kubeflow Pipelines to manage and run experiments, automatically track metrics, and use the Kubeflow Pipelines API for querying results.
E
Combine the use of AI Platform Training for experiment execution and Kubeflow Pipelines for workflow management, storing metrics in both Cloud Monitoring and BigQuery for comprehensive analysis.