Free Microsoft Azure AI Fundamentals MCQ Questions and Answers — Questions and Answers

Question 1: Which kind of artificial intelligence workload employs sensors to foretell future equipment mechanical problems and inform users beforehand?

Computer vision
Anomaly detection (Correct answer)
Conversational AI
Natural language processing

Correct answer: Anomaly detection

The task of using sensors to foretell future equipment mechanical problems and inform users beforehand is a classic application of anomaly detection. Anomaly detection AI workloads identify unusual patterns or outliers in data that deviate from normal behavior. These deviations can indicate impending failures or issues in machinery, allowing for proactive maintenance.

Question 2: The idea that responsible AI-based solutions have constraints is known as the _______________ principle.

Transparency (Correct answer)
Accountability
Privacy and security
Reliability and safety

Correct answer: Transparency

The principle that responsible AI-based solutions have constraints and that their limitations should be understood and communicated is a key aspect of Transparency. Transparency ensures that users and developers are aware of what an AI system can and cannot do, its potential biases, and the context in which it should be used, fostering realistic expectations and trust.

Question 3: The idea that everyone in society, regardless of gender or race, should benefit from ethical AI systems is known as: .

Accountability
Inclusiveness (Correct answer)
Privacy and security
Reliability and safety

Correct answer: Inclusiveness

The idea that everyone in society, regardless of gender or race, should benefit from ethical AI systems is the core definition of the Inclusiveness principle. This principle advocates for designing AI that is accessible, fair, and beneficial to the widest possible range of people. It aims to ensure no group is unfairly excluded or disadvantaged by AI technologies.

Question 4: Determine numerical values that, given a dog's age and body fat percentage, indicate the likelihood that it will get sick. What kind of machine learning model need to you employ?

Logistic regression
Linear regression
Hierarchical clustering
Multiple linear regression (Correct answer)

Correct answer: Multiple linear regression

The goal is to determine numerical values (likelihood of sickness) based on multiple input factors (dog's age and body fat percentage). Multiple linear regression is the appropriate machine learning model for predicting a continuous numerical outcome using two or more independent numerical features. Logistic regression is for binary classification, and hierarchical clustering is for grouping, not numerical prediction.

Question 5: Based on the dogs' age and body fat percentage, you want to utilize machine learning to estimate how sick the dogs will become. The model has to have .

three labels
three features
two features and one label (Correct answer)
two labels and one feature

Correct answer: two features and one label

In machine learning, features are the input variables used to make a prediction, and the label is the output variable that the model is trained to predict. Here, 'dog's age' and 'body fat percentage' are the two input features. The 'how sick the dogs will become' (the likelihood or degree of sickness) is the single numerical label that the model will learn to predict.

Question 6: You develop a model for multiclass classification. You must assess the model. Which two metrics for evaluation are available?

Accuracy (Correct answer)
F1 score (Correct answer)
Mean Squared Error (MSE)
Rand index

Correct answer: Accuracy

For multiclass classification models, Accuracy and F1 Score are common and appropriate evaluation metrics. Accuracy measures the proportion of correctly classified instances, while the F1 Score is the harmonic mean of precision and recall, providing a balanced measure that is particularly useful when class distributions are imbalanced. Mean Squared Error (MSE) is used for regression tasks, not classification.

Question 7: You prepare an Azure Machine Learning model for deployment as a predictive service in a real-world setting. Before deploying the model to _______________, an inference cluster needs to be created.

an Azure Logic Apps app
an Azure Function
Azure Container Instance
Azure Kubernetes Service (Correct answer)

Correct answer: Azure Kubernetes Service

When deploying a machine learning model as a predictive service in a real-world setting, especially for production-grade, scalable, and highly available inference, an inference cluster is required. Azure Kubernetes Service (AKS) is the recommended and most robust option within Azure Machine Learning for deploying models at scale, offering capabilities like auto-scaling, load balancing, and rolling updates for reliable production deployments.

Question 8: You want to use Azure Machine Learning's AutoML UI to create and deploy a predictive model. What are the three supported machine learning tasks?

Classification (Correct answer)
Forecasting (Correct answer)
Regression (Correct answer)
Reinforcement learning
Clustering

Correct answer: Classification

Azure Machine Learning's Automated Machine Learning (AutoML) UI is designed to simplify the process of building and deploying models for common supervised learning tasks. These tasks include Classification (predicting discrete categories), Regression (predicting continuous numerical values), and Forecasting (predicting future values based on time-series data). Reinforcement learning and clustering are different types of machine learning paradigms not directly supported by AutoML's core UI capabilities for predictive modeling.

Question 9: What method is the foundation for current picture categorization solutions?

Linear regression
Deep learning (Correct answer)
Multiple linear regression
Anomaly detection

Correct answer: Deep learning

Deep learning, particularly through the use of Convolutional Neural Networks (CNNs), is the foundational method for modern picture categorization solutions. These neural networks excel at automatically learning complex hierarchical features directly from raw image data, leading to state-of-the-art performance in tasks like image classification, object detection, and facial recognition. Linear regression and anomaly detection are not primary methods for general picture categorization.

Question 10: Using _____________________, you may extract the information from food product labels.

image segmentation
image classification
natural language processing
optical character recognition (Correct answer)

Correct answer: optical character recognition

Optical Character Recognition (OCR) is the technology specifically used to extract text from images or scanned documents, converting it into a machine-readable format. Food product labels contain printed text, such as ingredients, nutritional information, and brand names. Therefore, OCR is the appropriate method to digitize and extract this textual information for further processing or analysis.

Question 11: What are the Azure Computer Vision service's two features?

Data visualization
Model authoring
Model training (Correct answer)
Class prediction (Correct answer)

Correct answer: Model training

The Azure Computer Vision service offers capabilities for analyzing images, which includes performing 'Class prediction' by identifying objects, scenes, and actions within an image using pre-trained models. Additionally, through its Custom Vision sub-service, it allows users to upload their own images and labels to perform 'Model training' for custom image classification or object detection tasks tailored to specific needs. This enables the creation of specialized computer vision solutions.

Question 12: To find brand logos in photos, you must train a machine learning model. Which should you employ?

Azure Face Service
Azure Custom Vision object detection
Language Understanding Intelligent Service (LUIS)
Azure Custom Vision image classification (Correct answer)

Correct answer: Azure Custom Vision image classification

To train a machine learning model to identify brand logos in photos, you should employ Azure Custom Vision. This service allows you to build and deploy custom image recognition models. While 'object detection' within Custom Vision would be ideal for localizing logos with bounding boxes, 'image classification' can also be used to classify an image as containing a specific brand logo, thereby 'finding' its presence. Among the given options, Azure Custom Vision image classification is the most suitable for training a custom model to recognize specific visual patterns like logos.

Question 13: From a collection of PDF documents, you need to gather the names of individuals, groups, and occasions. Which natural language processing feature ought to be used?

Sentiment analysis
Extractive summarization
Key phrase extraction
Named entity recognition (Correct answer)

Correct answer: Named entity recognition

Named Entity Recognition (NER) is a natural language processing (NLP) feature specifically designed to identify and classify named entities in text into predefined categories. These categories typically include person names, organizations, locations, and temporal expressions. Therefore, to gather names of individuals, groups, and occasions from PDF documents, NER is the most appropriate feature to extract this structured information.

Question 14: Which of these attributes does the Azure Text Analytics service support?

Multilingual content
Confidence scoring
Chatbot integration
All of the above (Correct answer)

Correct answer: All of the above

The Azure Text Analytics service is a comprehensive natural language processing (NLP) offering that supports several key attributes. It can process 'Multilingual content' across various languages, provides 'Confidence scoring' for its results (e.g., sentiment scores, entity recognition confidence), and is designed for easy 'Chatbot integration' to enhance conversational AI applications. Thus, all the listed attributes are supported by the service.

Question 15: Based on their voice, users must be recognized. Which feature of the Azure Speech service should you employ?

Speaker recognition (Correct answer)
Conversation transcription
Pronunciation assessment
Language Understanding Intelligent Service (LUIS)

Correct answer: Speaker recognition

Speaker recognition is a specific feature within the Azure Speech service that identifies or verifies individuals based on their unique voice characteristics. This technology analyzes vocal patterns to distinguish between different speakers. Therefore, to recognize users based on their voice, speaker recognition is the appropriate feature to employ for tasks like voice authentication or personalization.

Question 16: By using ________________, you can swap chatbot activity with other services.

turns
dialog
channels (Correct answer)
cards

Correct answer: channels

In chatbot development, 'channels' refer to the various platforms or services through which users can interact with a bot. Examples include web chat, Microsoft Teams, Facebook Messenger, or Slack. By configuring different channels, a chatbot can integrate its activity and communicate with users across these diverse services, effectively 'swapping' its presence to reach a wider audience.

Free Microsoft Azure AI Fundamentals MCQ Questions and Answers

Which kind of artificial intelligence workload employs sensors to foretell future equipment mechanical problems and inform users beforehand?