Ultimate access to all questions.
Answer-first summary for fast verification
Answer: Utilize Dataproc for training your existing Spark ML models, reading data directly from BigQuery
The optimal choice is to use Dataproc for training your existing Spark ML models while directly accessing data from BigQuery. This approach avoids the unnecessary step of exporting data to a Spark cluster on Compute Engine. While rewriting models in TensorFlow and using Vertex AI is a viable option, sticking with Spark ML is more efficient given the existing setup. Option C efficiently combines the use of existing Spark ML models with the advantages of BigQuery.
Author: LeetQuiz Editorial Team
No comments yet.
You are an employee at an advertising company who has developed a Spark ML model to predict click-through rates for ads. With your company migrating to Google Cloud and your data being moved to BigQuery, you need to migrate your existing training pipelines that periodically retrain your Spark ML models. What is the most effective approach for this migration?
A
Rewrite your models using TensorFlow and utilize Vertex AI for training
B
Deploy a Spark cluster on Compute Engine and train Spark ML models with data exported from BigQuery
C
Utilize Dataproc for training your existing Spark ML models, reading data directly from BigQuery
D
Employ Vertex AI to train your existing Spark ML models