View Full Microsoft AI-900 Exam Dumps and Practice Test Dumps.

 

Question 1

What is the primary purpose of artificial intelligence?

1. To enable computers to perform tasks that normally require human intelligence
2. To replace all computer networks with cloud storage
3. To increase the physical memory of a computer
4. To create only mathematical calculations

Correct Answer: 1

Explanation

Artificial intelligence enables computer systems to perform tasks that typically require human-like intelligence. These tasks can include recognizing images, understanding language, making predictions, identifying patterns, and generating content. AI systems can use rules, machine learning models, or other techniques depending on the problem being solved. AI is not limited to one specific technology or application. It can be used in areas such as customer service, healthcare, finance, manufacturing, transportation, and many other industries where intelligent processing of information is useful.

Question 2

A company wants to predict whether a customer will cancel a subscription based on historical customer records. The historical records contain examples labeled as either “canceled” or “active.” Which machine learning task should be used?

1. Clustering
2. Classification
3. Regression
4. Anomaly detection

Correct Answer: 2

Explanation

Classification is used when a model predicts a category or class based on input data. In this scenario, the model must determine whether a customer belongs to the canceled or active category. Because historical examples already contain the expected outcomes, the model can learn the relationship between customer characteristics and the corresponding class. Regression would be used to predict a numerical value, while clustering groups unlabeled data based on similarities. Anomaly detection focuses on identifying unusual observations rather than assigning predefined categories.

Question 3

Which machine learning approach uses labeled training data to learn a relationship between inputs and known outputs?

1. Unsupervised learning
2. Reinforcement learning
3. Supervised learning
4. Generative AI

Correct Answer: 3

Explanation

Supervised learning uses training examples where the desired output or label is already known. The model learns patterns connecting the input features with those known outcomes and then uses those patterns to make predictions on new data. Classification and regression are common supervised learning tasks. In contrast, unsupervised learning works without predefined labels, while reinforcement learning learns through interactions and rewards. Generative AI focuses on producing new content based on learned patterns and is not simply another name for supervised learning.

Question 4

A retailer has customer purchasing data but does not have predefined customer categories. The retailer wants an AI system to discover groups of customers with similar purchasing behavior. Which technique is most appropriate?

1. Regression
2. Classification
3. Speech recognition
4. Clustering

Correct Answer: 4

Explanation

Clustering is an unsupervised learning technique used to organize data into groups based on similarities without requiring predefined labels. In this scenario, the retailer does not already know which customer categories should exist. A clustering algorithm can analyze purchasing characteristics and identify groups with similar behavior. These discovered groups could later help with marketing or customer analysis. Classification would require known categories, while regression predicts numerical values. Speech recognition is unrelated because the data in this scenario consists of customer purchasing information.

Question 5

A real estate company wants to predict the selling price of a house using its size, number of bedrooms, location, and age. Which machine learning task is appropriate?

1. Regression
2. Classification
3. Clustering
4. Object detection

Correct Answer: 1

Explanation

Regression is used when a machine learning model needs to predict a numerical value. A house selling price is a continuous numerical value, making regression suitable for this scenario. The model can use features such as house size, number of bedrooms, location, and age to learn relationships between these characteristics and historical selling prices. Classification would instead predict categories, clustering would discover groups without predefined labels, and object detection identifies objects and their locations within images. Regression is therefore the appropriate machine learning task for predicting house prices.

Question 6

A bank wants to identify unusual transactions that significantly differ from a customer’s normal spending behavior. Which AI capability is most appropriate?

1. Translation
2. Anomaly detection
3. Text summarization
4. Image classification

Correct Answer: 2

Explanation

Anomaly detection is designed to identify observations that differ significantly from expected patterns. In banking, this capability can help identify transactions that are unusual compared with a customer’s normal spending behavior. The system might consider factors such as transaction amount, location, timing, or purchasing patterns when determining whether an observation is unusual. Translation is used to convert content between languages, text summarization creates shorter versions of text, and image classification assigns categories to images. Therefore, anomaly detection is the best fit for this requirement.

Question 7

Which Azure service provides a collection of prebuilt AI capabilities for tasks such as language processing, speech, and computer vision?

1. Azure Virtual Machines
2. Azure Storage
3. Azure AI services
4. Azure DNS

Correct Answer: 3

Explanation

Azure AI services provide ready-to-use artificial intelligence capabilities that developers can integrate into applications without building every AI model from the beginning. These services include functionality for areas such as computer vision, speech processing, and language understanding. This makes them useful when an organization needs common AI capabilities quickly. Azure Virtual Machines provide compute resources, Azure Storage provides cloud storage capabilities, and Azure DNS handles domain name resolution. Azure AI services are specifically focused on providing managed AI capabilities for application development.

Question 8

A manufacturing company wants to automatically inspect product photographs and determine whether each product contains a visible defect. Which AI workload is most appropriate?

1. Natural language processing
2. Speech synthesis
3. Computer vision
4. Reinforcement learning

Correct Answer: 3

Explanation

Computer vision enables AI systems to analyze and interpret visual information such as photographs and video. In a manufacturing environment, computer vision can be used to inspect products and identify visible defects, classify images, detect objects, or analyze other visual characteristics. Natural language processing works with human language, while speech synthesis converts text into spoken audio. Reinforcement learning focuses on learning actions through rewards and feedback. Because the manufacturing requirement involves analyzing product photographs, computer vision is the appropriate AI workload.

Question 9

An application needs to identify multiple cars, people, and bicycles in a street photograph and determine where each object appears in the image. Which computer vision capability is required?

1. Object detection
2. Image classification
3. Text translation
4. Sentiment analysis

Correct Answer: 1

Explanation

Object detection identifies objects within an image and determines their locations, typically by providing bounding boxes around detected objects. In a street photograph, an object detection system could identify several cars, people, and bicycles and indicate where each one appears. Image classification generally assigns one or more categories to an entire image rather than locating individual objects. Text translation converts content between languages, while sentiment analysis evaluates the emotional tone of text. Object detection is therefore the appropriate computer vision capability for this requirement.

Question 10

A company analyzes customer reviews to determine whether customers are expressing positive, negative, or neutral opinions. Which AI capability should it use?

1. OCR
2. Object detection
3. Speech recognition
4. Sentiment analysis

Correct Answer: 4

Explanation

Sentiment analysis evaluates text to determine the expressed sentiment, commonly categorized as positive, negative, or neutral. Businesses can use this capability to analyze customer reviews, survey responses, social media comments, or support messages and understand overall customer reactions. OCR is used to extract text from images and documents, object detection identifies objects in images, and speech recognition converts spoken language into text. Since the company wants to determine the emotional tone of written customer reviews, sentiment analysis is the appropriate AI capability.

Question 11

Which AI workload focuses on enabling computers to understand, interpret, and process human language?

1. Natural language processing
2. Computer vision
3. Anomaly detection
4. Regression

Correct Answer: 1

Explanation

Natural language processing, or NLP, focuses on enabling computer systems to work with human language. NLP capabilities can include analyzing text, identifying entities, extracting key phrases, translating languages, summarizing documents, and determining sentiment. Computer vision instead processes visual information such as images and video. Anomaly detection identifies unusual patterns, while regression predicts numerical values. NLP is widely used in applications such as chatbots, search systems, document processing, virtual assistants, and customer feedback analysis because these solutions need to process or understand human language.

Question 12

A voice assistant receives a spoken question from a user and needs to convert the speech into written text before processing the question. Which capability is required?

1. Text-to-speech
2. Speech-to-text
3. Image classification
4. Key phrase extraction

Correct Answer: 2

Explanation

Speech-to-text converts spoken language into written text. In a voice assistant, this capability can transform the user’s spoken question into text so that other AI components can process and interpret it. After transcription, the application might use natural language processing to understand the request and generate an appropriate response. Text-to-speech performs the opposite operation by converting written text into spoken audio. Image classification analyzes images, while key phrase extraction identifies important terms from existing text. Speech-to-text is therefore the necessary first step in this scenario.

Question 13

An organization wants to ensure that an AI model does not consistently produce worse results for one demographic group than for another. Which responsible AI principle is most directly related to this requirement?

1. Transparency
2. Privacy
3. Fairness
4. Reliability

Correct Answer: 3

Explanation

Fairness focuses on ensuring that AI systems treat people equitably and do not produce unjustified differences in outcomes for different groups. Organizations can evaluate model performance across relevant populations to identify potential disparities in accuracy, error rates, or other outcomes. Transparency concerns making AI behavior and decision-making understandable, while privacy focuses on protecting personal information. Reliability concerns consistent and dependable operation. When the main concern is whether different demographic groups experience unequal model performance or treatment, fairness is the most directly relevant responsible AI principle.

Question 14

A company wants users to understand why an AI system produced a particular recommendation and what factors influenced its output. Which responsible AI principle is most relevant?

1. Inclusiveness
2. Transparency
3. Privacy
4. Security

Correct Answer: 2

Explanation

Transparency involves making information about an AI system understandable to relevant users and stakeholders. This can include explaining how the system works, what data or factors may influence its outputs, and what limitations users should understand. Transparency can improve trust and help people make informed decisions when interacting with AI systems. Inclusiveness focuses on designing systems that work for people with different needs and abilities. Privacy protects personal information, while security helps protect systems and data from threats. The requirement described is primarily related to transparency.

Question 15

An AI application processes customers’ personal information. Which practice is most important for protecting that information?

1. Collecting every available data field
2. Making all customer data publicly accessible
3. Ignoring data access controls
4. Applying appropriate privacy and security controls

Correct Answer: 4

Explanation

Privacy and security controls help protect personal information from inappropriate access, use, disclosure, or loss. Organizations should consider practices such as limiting data collection to what is necessary, controlling access, protecting stored and transmitted information, and following applicable privacy requirements. Collecting unnecessary information can increase risk, while publicly exposing customer data or ignoring access controls creates significant security concerns. Responsible AI systems should account for privacy and security throughout the solution lifecycle, especially when sensitive or personal information is processed.

Question 16

An AI system used in an important business process should continue operating consistently and safely when it encounters unexpected conditions. Which responsible AI principle is most closely associated with this requirement?

1. Reliability and safety
2. Inclusiveness
3. Transparency
4. Fairness

Correct Answer: 1

Explanation

Reliability and safety focus on ensuring that an AI system performs consistently and safely under expected conditions and can appropriately handle situations that may cause problems. Important AI applications should be tested for failures, unexpected inputs, and operational risks before and during deployment. Inclusiveness focuses on serving people with different needs, transparency focuses on understandable behavior and information, and fairness addresses equitable treatment across groups. When the primary concern is dependable and safe operation, reliability and safety are the most relevant responsible AI principles.

Question 17

Which type of AI can create new content such as text, images, audio, or code based on a user’s prompt?

1. Classification AI
2. Generative AI
3. Clustering AI
4. Regression AI

Correct Answer: 2

Explanation

Generative AI is designed to create new content based on patterns learned from data. Depending on the model and application, it can generate text, images, audio, code, and other types of content. Users commonly interact with generative AI by providing prompts that describe what they want the system to produce. Classification assigns inputs to categories, clustering discovers groups within data, and regression predicts numerical values. Generative AI therefore differs from traditional predictive machine learning because its primary purpose includes producing new content rather than only predicting an existing outcome.

Question 18

A retailer wants to predict the amount of money each customer is likely to spend during the next month. The expected result is a numerical amount. Which machine learning task should be selected?

1. Classification
2. Clustering
3. Regression
4. Object detection

Correct Answer: 3

Explanation

Regression is appropriate when the desired prediction is a numerical value. In this scenario, the model could learn from historical customer information such as previous purchases, transaction frequency, and other relevant features to estimate the amount a customer may spend during the next month. Classification would produce a category rather than a continuous numerical amount. Clustering would discover groups without predefined target values, and object detection is a computer vision capability. Because the expected output is a monetary amount, regression is the appropriate machine learning task.

Question 19

A company wants to build a chatbot that can understand customer questions written in natural language and provide appropriate responses. Which AI workload is most relevant?

1. Computer vision
2. Natural language processing
3. Anomaly detection
4. Image classification

Correct Answer: 2

Explanation

Natural language processing is central to chatbots that need to understand and respond to human language. NLP capabilities can help identify the meaning of user questions, extract relevant information, determine intent, and support the generation of appropriate responses. A conversational AI solution may combine NLP with other components such as search, knowledge bases, or generative AI. Computer vision and image classification focus on visual information, while anomaly detection identifies unusual patterns. Since the chatbot must understand written customer questions, NLP is the most relevant AI workload.

Question 20

Which statement best describes the relationship between machine learning and artificial intelligence?

1. Artificial intelligence is a subset of machine learning
2. Machine learning and artificial intelligence are completely unrelated
3. Machine learning is a subset of artificial intelligence
4. Machine learning can only be used for image processing

Correct Answer: 3

Explanation

Machine learning is a subset of artificial intelligence that enables systems to learn patterns from data and use those patterns to make predictions or decisions. Artificial intelligence is a broader field that includes machine learning as well as other approaches for creating systems capable of performing tasks associated with intelligence. Machine learning can be applied to many areas, including language, images, forecasting, recommendations, and fraud detection. Therefore, machine learning should be understood as one important approach within the broader field of artificial intelligence.