0% found this document useful (0 votes)
29 views10 pages

Mod 2

1. The document describes different Azure services and resources for natural language processing, computer vision, and machine learning. It includes frequently asked questions about how to use these services for different use cases. 2. Recommendations are provided on which Azure services and resources to use for tasks such as predicting sales, classifying images, detecting defects, predicting used car prices, classifying loan applications, and more. 3. It also includes tips on how to set up and evaluate.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
29 views10 pages

Mod 2

1. The document describes different Azure services and resources for natural language processing, computer vision, and machine learning. It includes frequently asked questions about how to use these services for different use cases. 2. Recommendations are provided on which Azure services and resources to use for tasks such as predicting sales, classifying images, detecting defects, predicting used car prices, classifying loan applications, and more. 3. It also includes tips on how to set up and evaluate.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

1.

He wants to create a model to predict ice cream sales based on data.


historical data that includes daily total sales and weather measurements. What
Which Azure service should be used?

Azure Machine Learning

2.

Wants to train a model that classifies images of dogs and cats based on
from a collection of your own digital photographs. What Azure service should
to use?

Custom Vision

3.

He/She is going to design an artificial intelligence application that uses computer vision to
detect cracks in automobile windshields and alerts drivers
when to repair or replace a windshield. When it is tested in good conditions
lighting conditions, the application correctly detects 99% of the
damaged glasses. Which of the following statements should be included in the interface?
app user?

When used under good lighting conditions, this application can be used.
to identify possible cracks and potentially hazardous defects in windshields. If
suspects that the windshield is damaged, even if the application does not detect any defect,
It must be inspected by a professional.

1. 2

A dealership wants to use historical car sales data to train


a Machine Learning model. The model must predict the price of a
second-hand car depending on the brand, model, size of the
engine and the mileage. What type of Machine Learning model should be created?
dealership with automated machine learning?

Regression
2.

A bank wants to use historical records of loan defaults to


classify loan requests as low risk or high risk, in
characteristics function such as the amount or term of the loan and the income
from the borrower. What type of Machine Learning model should the bank create?
with automated machine learning?

classification

3.

He wants to use automated machine learning to train a model of


regression with the best possible R2 score. How should I set up the
automated machine learning experiment?

Establish the main metric in the R2 score

1. 3

You are creating a training pipeline for a regression model. It


use a dataset that has several numerical columns in which the
values are on different scales. Wants to transform the columns
numerical so that all values are on a similar scale. It also wants
that the transformation is scaled in relation to the minimum and maximum values of
Each column. What module should be added to the pipeline?

Data normalization

2.

Why do you split the data into training and validation sets?

Dividing the data into two sets allows for comparing the labels predicted by the model
with the known real labels of the original dataset.
1. 4

You are going to use the Azure Machine Learning designer to create a pipeline.
training for a binary classification model. A set has been added
data that contains features and labels, a Decision Tree module
of two classes and a module Train model. It is planned to use the modules
Score model and Evaluate model to test the trained model with a
subset of the dataset that has not been used for training.
What kind of additional module should be added?

Split Data

2.

Use an Azure Machine Learning designer pipeline to train and


test a binary classification model. Review the performance metrics of the
model in a module Evaluate model and notice that it has an AUC score
From 0.3. What conclusion can be drawn from the model?

The model is worse than random estimation.

3.

Use the Azure Machine Learning designer to create a pipeline


training for a classification model. What should you do before
implement the model as a service?

Create an inference pipeline from the training pipeline

You are using an Azure Machine Learning designer pipeline to


train and test a K-means clustering model. He wants that
model assign elements to one of the three clusters. What property of the
The configuration of the K-means clustering module must be set
to achieve it?

Set the number of centroids to 3


2.

Use the Azure Machine Learning designer to create a pipeline.


training for a clustering model. Now wants to use the
model in an inference pipeline. What module should be used to deduce the
cluster model predictions?

Data assignment to clusters

1. 6

He wants to use the Computer Vision service to analyze images. He also wishes to
use the Text Analytics service to analyze the text. The goal is for the
developers only need a key and an endpoint to
access all services. What type of resource should we create in the subscription?
from Azure?

Cognitive Services

2.

He wants to use the Computer Vision service to identify the location of the
individual elements of an image. Which of the following characteristics must
recover?

Objects

3.

Do you want to use the Computer Vision service to analyze images of locations and
identify well-known buildings. What should we do?

Retrieve the categories of the image, specifying the domain of reference points.
1. 7

Are you planning to use the Custom Vision service to train a solution model?
image classification. He wants to create a resource that can only be used for the
model training and not for prediction. What type of resource should be created?
in the Azure subscription?

Custom Vision

2.

Train an image classification model that achieves evaluation metrics


that are not completely satisfactory. How could I improve it?

By adding more images to the training set.

3.

He has published an image classification model. What information should


provide to developers who want to use it?

The project ID
of prediction.

1. 8

Which of the following results makes an object detection model


Does it normally return an image?

A class label, a probability, and a bounding box for each object in the image

2.

He plans to use a set of images to train a model of


object detection and then publish it as a predictive service. He wants
use a single Azure resource with the same key and the same endpoint
for training and prediction. What type of Azure resource should be created?

Cognitive Services
1. 9

It is planned to use the face to detect human faces in an image. How


indicates the service the location of the faces it detects?

A set of coordinates for each of the faces, which defines a bounding box
rectangular around the surface.

2.

Which of the following aspects can affect facial detection?

Extreme angles

3.

He wants to use Face to identify users by name. What should he do?

Use Face to create a group that contains several images of each individual with their name
and train a model based on the group

1. 10

We want to extract text from images and then use the Text Analytics service to
analyze the text. Our goal is for developers to only need
a key and a connection point to access all services. What type of
What resource should we create in the Azure subscription?

Cognitive Services

2.

Do you plan to use the Computer Vision service to read text in a document?
Large PDF. Which API should I use?

Read API
1. 11

He plans to use the pre-generated receipt model from Form Recognizer. What
What type of Azure resource should you create?

Form Recognizer resource or Cognitive Services

2.

Use the Form Recognizer service to analyze receipts that you have scanned.
Images in JPG format. What is the maximum file size of the JPG file?
What can you send to the pre-generated receipt model?

50 MB

12

He wants to use the Language service to determine the key talking points.
in a text document. What feature of the service should I use?

Extraction of key phrases

2.

You are using the Language service to conduct sentiment analysis on a


document and this returns a score of 0.99. What does this score indicate?
about the opinion of the document?

The document is positive.

3.

When is it possible to see a score with a NaN value Detection of


language?

When the language is ambiguous.


1. 13

He plans to compile an application that uses the Voice service for transcription.
audio recordings of phone calls and then sends the transcribed text to
Text Analytics service to extract key phrases. You are interested in managing access.
and the billing of application services in a single Azure resource. What
What type of Azure resource should you create?

Cognitive Services

2.

He wants to use the Voice service to compile an application that reads out loud the
incoming email message issues. What API should be used?

Text-to-Speech

14

You are developing an application that must take input in English from a
microphone and generate a real-time text-based transcription in Hindi.
What service should he/she use?

Speech

2.

You must use the Translator Text service to translate email messages.
What is the most effective way to achieve it?

Make a single call to the service, specifying a 'source' language of 'es', a


idioma de "destino" de "en" y otro idioma de "destino" de "fr".

15

You must provision an Azure resource that will be used to create an application.
Language Understanding. What type of resource should you create?

Language service
2.

You are going to create a conversational language recognition application for


to admit an international clock. He wants users to be able to request the time
current in a specific city, for example, "What time is it in London?" What
What do you have to do?

city
intention of the city.

3.

It has published the conversational language recognition application. What


information a client application developer needs to obtain
her predictions?

The connection point and the key of the application's prediction resource.

1. 16

Your organization has a frequently asked questions (FAQ) document. It is


It is necessary to create a knowledge base that includes the questions and answers of the
Frequently asked questions with the least effort possible. What do you have to do?

Import the existing frequently asked questions document into a new database.
knowledge.

2.

You must provide a technical support bot for internal use in your
organization. Some users want to be able to send questions to the bot with
Microsoft Teams, while others want to use a web chat interface in a
internal website. What does it have to do?

Create a knowledge base. Next, create a bot for the knowledge base.
and connect the web chat channels and Microsoft Teams for your bot.
1. 17

Your organization has a frequently asked questions (FAQ) document.


It is necessary to create a knowledge base that includes the questions and answers of the
Frequently asked questions with the least effort possible. What do you have to do?

Import the existing frequently asked questions document to a new database.


knowledge.

2.

You must provide a technical support bot for internal use in your
organization. Some users want to be able to send questions to the bot with
Microsoft Teams, while others want to use a web chat interface in a
internal website. What does it have to do?

Create a knowledge base. Next, create a bot for the knowledge base.
and connect the web chat and Microsoft Teams channels for your bot.

You might also like