Answer-first summary for fast verification
Answer: Automatic deployment and serving, Hyperparameter tuning
BigQuery ML is designed to simplify the machine learning process by allowing users to create and execute ML models using SQL queries within BigQuery. While it supports a range of functionalities like exploratory data analysis (through SQL queries), feature selection (manually via SQL), model building, and training, it does not natively support automatic deployment and serving of models. Additionally, while it offers basic hyperparameter tuning for certain models, it lacks advanced or automated tuning features. For deployment and advanced tuning, integrating with other Google Cloud services like AI Platform Prediction is recommended. This makes options F and E the correct choices as they highlight the limitations of BigQuery ML in these areas.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
Your company, which lacks extensive machine learning (ML) expertise, is exploring BigQuery ML as a managed service to initiate its ML journey. The goal is to leverage a service that is straightforward and requires minimal overhead. Given this scenario, and considering the need for a comprehensive ML workflow from data exploration to model deployment, which of the following features does BigQuery ML not natively support? Choose the two most correct options.
A
Exploratory data analysis
B
Feature selection
C
Model building
D
Training
E
Hyperparameter tuning
F
Automatic deployment and serving
No comments yet.