Exam Cloud Digital Leader topic 1 question 248 discussion

Actual exam question from Google's Cloud Digital Leader

Question #: 248
Topic #: 1

A real estate organization processes photos of properties uploaded by their agents to an internal application. They must manually check and remove photos that contain personally identifiable information like passports and credit cards. They want a solution that their developer can use to integrate with their app as quickly as possible. Which solution should they use?

A. Pre-trained APIs
B. AutoML
C. Custom training
D. BigQuery ML

Show Suggested Answer

Suggested Answer: A 🗳️

by Moin23 at Sept. 19, 2024, 11:52 a.m.

Comments

Submit Cancel

joshnort

6 months, 2 weeks ago

Selected Answer: A

A. Pre-trained APIs. Pre-trained APIs are ready-to-use models provided by Google Cloud, such as the Cloud Vision API, which can detect a wide range of sensitive information, including personally identifiable information (PII) like passports, credit cards, and other sensitive data in images. These APIs are quick to integrate into applications, as they don't require additional training, allowing the developer to focus on the application rather than building or fine-tuning models.

upvoted 1 times

joshnort

6 months, 2 weeks ago

Why the other options are incorrect: B. AutoML: While AutoML allows developers to build custom models with machine learning, it requires more effort than pre-trained APIs. It involves training a model on custom data, which might be beneficial if the organization needs a model tailored to their specific use case. However, it’s not as quick or straightforward as using a pre-trained API, which is what the organization needs for fast integration. C. Custom training: Custom training refers to the process of creating entirely custom machine learning models from scratch. This option requires a deep understanding of machine learning and more time to prepare and train the model. Given the organization's need for a fast solution, this approach would be overkill and not as efficient as using pre-trained APIs.

upvoted 1 times

joshnort

6 months, 2 weeks ago

D. BigQuery ML: BigQuery ML allows users to build machine learning models directly within BigQuery, typically for structured data like tabular data. It is not well-suited for image processing, which is the need here (analyzing photos for PII). This option is more appropriate for data analytics or predictions based on structured data rather than image-based PII detection.

upvoted 1 times

...

Ekramy_Elnaggar

8 months ago

Selected Answer: A

Cloud Vision API

upvoted 2 times

...

Moin23

10 months, 2 weeks ago

Selected Answer: A

Answer A

upvoted 2 times

...