Google Professional Machine Learning Engineer Question 241
Select 2Google Cloud PlatformYou are designing a data pipeline to ingest and preprocess datasets of various types including CSV files, JSON files, images, and relational database records for training a machine learning model. Which of the following tools or services in Google Cloud can you use to efficiently ingest this data into a centralized storage system for further processing?
- A
Cloud Storage
- B
BigQuery
- C
Cloud Dataflow
- D
TensorFlow Extended (TFX)
- E
Cloud SQL
Show answer and explanation
Correct answers: A, C
Explanation
To ingest and preprocess diverse file types like CSV, JSON, and images, Cloud Storage is an ideal choice for centralized storage, while Cloud Dataflow can be used for data transformation and ingestion into Cloud Storage or other systems. BigQuery, TFX, and Cloud SQL are not designed for these ingestion tasks but serve other purposes in data workflows.
- A. Correct.
Cloud Storage is a highly scalable and cost-effective solution for ingesting and storing various types of data such as CSV, JSON, images, and other file formats. It is suitable for centralized storage of raw data.
- B. Incorrect.
BigQuery is mainly used for querying and analyzing structured data but is not a direct ingestion tool for diverse file types. It works well after data has been preprocessed and structured.
- C. Correct.
Cloud Dataflow is a fully managed service for stream and batch data processing. It can handle the transformation and ingestion of various data types into centralized systems like Cloud Storage or BigQuery.
- D. Incorrect.
TensorFlow Extended (TFX) is a machine learning pipeline tool focused on model development and deployment, not on raw data ingestion or storage.
- E. Incorrect.
Cloud SQL is a managed relational database service best suited for structured database workloads, but it is not designed for ingesting diverse file types into a centralized storage system.