Google Associate Cloud Engineer Question 141
Single answerGoogle Cloud PlatformYour company is planning to deploy a machine learning model as a data product using Google Cloud Platform. The model needs to be accessible via a REST API and should automatically scale based on traffic. Which Google Cloud service should you use to host this model?
- A
Google Cloud Storage
- B
Google Cloud Functions
- C
Google Kubernetes Engine
- D
AI Platform Prediction
Show answer and explanation
Correct answer: D
Explanation
AI Platform Prediction is the most suitable service for deploying machine learning models on Google Cloud Platform, providing a scalable REST API endpoint for model inference. It is specifically designed to handle model serving and abstracts the complexities of infrastructure management, making it an ideal choice for this scenario.
- A. Incorrect.
Google Cloud Storage is used for storing data objects but does not offer the capability to host and serve machine learning models as a REST API.
- B. Incorrect.
Google Cloud Functions can run code in response to events and can be used to serve simple models, but it is not optimized for deploying complex machine learning models that require scaling based on traffic.
- C. Incorrect.
Google Kubernetes Engine can be used to deploy machine learning models, but it requires managing the Kubernetes clusters and is not specifically optimized for model serving.
- D. Correct.
AI Platform Prediction is specifically designed for deploying machine learning models. It provides a REST API and can automatically scale the model deployments based on the incoming traffic.