Which of the following is part of the azure artificial intelligence service?

Here are the correct answers with explanation:

1. b.
   Azure Bot Service will not help with prediction. It should be eliminated as a candidate.
2. a.
   Azure Cognitive Services includes Vision services that can identify the content of an image. Azure Cognitive Services is the best candidate.
3. b.
   Azure Pipelines is a CI/CD tool for building an automated toolchain. It lacks features to assign tasks for individual developers to work on. However, it can automate other tools to assign tasks to users.
4. c.
   Azure Service Health provides incident history and RCAs to share with your stakeholders.
5. d.
   The Azure CLI enables you to use Bash to run one-off tasks on Azure.
6. c.
   Azure Sphere provides the highest degree of security to ensure the device has not been tampered with.
7. b.
   Azure Logic Apps is best suited for users who are more comfortable in a visual environment that allows them to automate their business processes. Logic Apps is the best option in this scenario.
8. b.
   Azure Logic Apps makes it easy to create a workflow across well-known services with less effort than writing code and manually orchestrating all the steps yourself.
9. b.
   IoT Central quickly creates a web-based management portal to enable reporting and communication with IoT devices.
10. Although Azure Machine Learning could be used to create a natural language model, it would likely be cost and time prohibitive. It should be eliminated as a candidate.

Please share your feedback in the below comments section.

Check out the TestPrep material for more exam preparation material.

Disclaimer: These questions are NOT appearing in the certification exam. I personally or CloudThat do not have any official tie-up with Microsoft regarding the certification or the kind of questions asked. These are my best guesses for the kind of questions to expect with Microsoft in general and with the examination.

Feel free to drop any questions in the comment box, I would love to address them. I hope you enjoyed the article. Best of luck!

Pankaj JainaniFollowMay 3, 2020·5 min readSaveAzure Machine Learning Service: Part 1 — An Introduction (adsbygoogle = window.adsbygoogle || []).push({}); A starter series to develop & deploy end-to-end AI solutions on Microsoft Azure public cloud platform.

Graphic Credit: Microsoft for Azure Machine LearningPreface

Ever since I have successfully cleared my Microsoft’s DP-100 exam certification, which is “Designing and Implementing a Data Science Solution on Azure”. I was quite keen to write something in the form of KT so that it can help somebody someday. Hence, while keeping the motto in mind I am herewith the first article on the Azure Machine Learning Service (Azure ML Service). Since this is the first article, I am going to cover a very basic introduction to this completely cloud-managed service. Later, I will publish more articles on advanced topics, hence this is going to be a series.

Please note, for those who are having prior knowledge of Azure and fundamentals will find this and series of upcoming articles quite intuitive, for rest all others who are having limited knowledge about Azure will find this as a knowledge source with very high level overview of Cloud aspects of AI/ML development and deployment. Also, here are few resources to get started with this service from Microsoft’s official documentation page.

Pankaj Jainani - Cloud Solutions Architect - LinkedIn

I am a Cloud Solution Architect, Artificial Intelligence Engineer, DevOps Consultant, Scrum Master, Technical Manager…

www.linkedin.com

Introduction

Azure Machine Learning (Azure ML) is a cloud-based service for creating and managing machine learning solutions. It’s designed to help data scientists and machine learning engineers leverage their existing data processing and model development skills & frameworks. Also, help them to scale, distribute, and deploy their workloads to the cloud. The Azure ML SDK for Python provides classes that we can use to work with Azure ML in our Azure subscription.

Graphic Credit: Microsoft Azure ML Service Documentation

The entire writeup is adapted from my own Kaggle notebook — hosted here. So, if anyone like to directly jump into the working code please follow the same link.

Getting Started

Workspace for Azure ML resources

Workspace is a logical container for all AML assets: Compute, Storage, Data, Scripts, Notebooks, Experiments (their versions), Metrics, Pipelines, and Models (their versions).

The diagram below from Azure ML describes the logical Workspace:

Graphic Credit: Microsoft Azure ML Service Documentation

Once the workspace is created from either Azure Portal or AML SDK, we have to connect to the workspace to perform our machine learning operations. Hence, we need to start by importing the following packages from the SDK—

Import packages and libs

After all the packages are loaded in the notebook, we have to instantiate the workspace object ws using Workspace.get() method by passing value of AML service name, name of Azure resource group, and Azure subscription ID of AML service.

There is an alternate approach to instantiate the Workspace object using config.json file. This approach is explained in the notebook.

Experiment

The AML service Experiment is a named process, usually running of a script or a pipeline, that can generate metrics and outputs and be tracked in the Azure Machine Learning workspace. We can run the same experiment multiple times with different data, codes, or settings. The experiment is defined below:

Define and Start an Experiment

Once the experiment is submitted, run context is used to initialize, track, and complete the experiment. Within the run context we can capture various metrics from the experiment’s execution, these metrics are associated with its run object.

Using the Run Context to capture metrics from the Experiment’s run.

In the code above, we can capture a single value using run.log(), capture a pictorial representation of a metric image using run.log_image() and similarly, a list of values are captured using run.log_list() method.

Experiment Details

After the experiment is complete, the run object is used to get information and details for the instance of the experiment:

Grab the details of the experiment

These statements will result in an output related to the experiment, metadata, metrics, and logs.

Experiment Dashboard

Following are the dashboards readily available in Azure Portal for each experiment’s run:

1. The Name of the experiment is “simple-experiment”
2. Gives the details of each run of this experiment

Experiment’s Dashboard

Experiment’s DetailsWhat Next?

This is a series of blog posts encompassing a detailed overview of various Azure Machine Learning capabilities, the URLs for other posts are as follows:

Which of the following services are part of artificial intelligence service in Azure?

What are Azure AI services?

Which Azure tool can help you build artificial intelligence AI applications?

Which of the following are the services provided by Azure?