Exam AI-100 topic 4 question 19 discussion

Actual exam question from Microsoft's AI-100

Question #: 19
Topic #: 4

Note: This question is part of a series of questions that present the same scenario. Each question in the series contains a unique solution that might meet the stated goals. Some question sets might have more than one correct solution, while others might not have a correct solution.
After you answer a question, you will NOT be able to return to it. As a result, these questions will not appear in the review screen.
You have Azure IoT Edge devices that generate streaming data.
On the devices, you need to detect anomalies in the data by using Azure Machine Learning models. Once an anomaly is detected, the devices must add information about the anomaly to the Azure IoT Hub stream.
Solution: You expose a Machine Learning model as an Azure web service.
Does this meet the goal?

A. Yes
B. No

Show Suggested Answer

Suggested Answer: B 🗳️
Instead use Azure Stream Analytics and REST API.
Note. Available in both the cloud and Azure IoT Edge, Azure Stream Analytics offers built-in machine learning based anomaly detection capabilities that can be used to monitor the two most commonly occurring anomalies: temporary and persistent.
Stream Analytics supports user-defined functions, via REST API, that call out to Azure Machine Learning endpoints.
References:
https://docs.microsoft.com/en-us/azure/stream-analytics/stream-analytics-machine-learning-anomaly-detection

by T0p1cs at Sept. 28, 2019, 1:49 a.m.

Comments

Submit Cancel

T0p1cs

Highly Voted 5 years, 7 months ago

This is also a Yes, Machine Learning modules on the Edge can also publish data to IoT Hub. See the following tutorial. https://docs.microsoft.com/en-us/azure/iot-edge/tutorial-deploy-machine-learning

upvoted 7 times

...

ari_ira

Highly Voted 5 years, 5 months ago

I think an Azure ML model can be deployed as an IoT Edge Model, not exposed as an Azure Web service

upvoted 7 times

...

rveney

Most Recent 1 year, 10 months ago

B. No The solution of exposing a Machine Learning model as an Azure web service does not meet the goal of detecting anomalies in the data generated by Azure IoT Edge devices and adding information about the anomaly to the Azure IoT Hub stream. Exposing a Machine Learning model as an Azure web service would typically involve deploying the model as a web API endpoint that can be accessed by external applications or services. However, this approach does not address the requirement of running the model directly on the IoT Edge devices and adding anomaly information to the Azure IoT Hub stream. To achieve the goal, you would need to deploy an Azure Machine Learning model as an IoT Edge module. This allows you to run the machine learning model directly on the edge devices, enabling real-time anomaly detection on the devices themselves. Once an anomaly is detected, the IoT Edge module can add relevant information about the anomaly to the Azure IoT Hub stream.

upvoted 1 times

...

sayak17

4 years, 7 months ago

if by "expose" they also mean "deploy" then it is a Yes Proof: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-deploy-and-where?tabs=azcli

upvoted 2 times

Cornholioz

4 years, 2 months ago

But it says "add information about the anomaly to the Azure IoT Hub stream". Can deploying the ML do this? That's a job for Functions or ASA.

upvoted 1 times

...

Daniel_vahid

5 years, 3 months ago

Azure ML deployment to IoT edge is in preview yet.