You need to predict the price of a house based on its location and size. Which type of machine learning model should you use?
- Classification
- Regression (Correct)
- Clustering
- Anomaly detection
Explanation: Regression is a supervised machine learning technique used to predict continuous numeric values, such as prices, temperatures, or quantities. Classification predicts discrete class labels, and clustering groups unlabeled data.
Which Azure service can be used to extract printed and handwritten text from images?
- Azure AI Language
- Azure AI Search
- Azure AI Vision (Correct)
- Azure AI Bot Service
Explanation: Azure AI Vision includes Optical Character Recognition (OCR) capabilities to detect and extract printed or handwritten text embedded within images and documents.
Which two components are required to build a conversational bot using Azure AI? (Choose two.)
- Azure AI Language (Correct)
- Azure AI Face
- Azure AI Bot Service (Correct)
- Azure AI Translator
Explanation: Building an Azure conversational bot typically requires a knowledge base or natural language model created in Azure AI Language (such as question answering) to understand and resolve user queries, combined with Azure AI Bot Service to publish channels and manage interactions.
You are designing an AI system that must treat all people fairly regardless of gender, race, or ethnicity. Which core principle of Responsible AI does this represent?
- Reliability and safety
- Fairness (Correct)
- Privacy and security
- Inclusiveness
Explanation: Fairness ensures AI systems treat all people impartially without discriminatory bias based on demographic attributes.
Which type of machine learning groups similar items together without using predefined labels?
- Classification
- Regression
- Clustering (Correct)
- Supervised learning
Explanation: Clustering is an unsupervised machine learning method that groups observations into clusters based on natural feature similarities without requiring prior class labels.
You want to identify commercial brand logos present in a collection of photographs. Which Azure AI service includes prebuilt support for this capability?
- Azure AI Custom Vision
- Azure AI Vision (Correct)
- Azure AI Face
- Azure AI Document Intelligence
Explanation: Azure AI Vision provides built-in brand detection to identify recognized commercial brand logos within photos.
Which two tasks are examples of natural language processing (NLP)? (Choose two.)
- Extracting key phrases from an article (Correct)
- Determining the sentiment of a product review (Correct)
- Identifying the location of a car in an image
- Predicting the future stock price of a company
Explanation: Natural Language Processing processes written or spoken human language. Extracting key phrases and evaluating sentiment are core NLP tasks.
Which feature of Azure Machine Learning provides a visual drag-and-drop canvas to build and deploy ML pipelines?
- Automated ML
- Notebooks
- Compute clusters
- Designer (Correct)
Explanation: Azure Machine Learning designer provides a visual interface to connect datasets and pre-built processing modules on a drag-and-drop canvas.
What is the primary objective of anomaly detection in machine learning workloads?
- To identify unusual patterns that do not conform to expected behavior (Correct)
- To predict continuous numerical values
- To group similar data points together into clusters
- To classify images into predefined categories
Explanation: Anomaly detection identifies rare events, deviations, or outliers that deviate significantly from typical historical patterns.
Which Responsible AI principle requires that AI systems perform reliably and safely even when encountering edge cases or hostile conditions?
- Fairness
- Inclusiveness
- Reliability and safety (Correct)
- Transparency
Explanation: Reliability and safety requires AI solutions to undergo rigorous testing to ensure consistent, secure, and predictable performance under unexpected operational conditions.
You want to train a computer vision model using your own dataset of specialized animal breeds. Which service allows you to train and publish custom classifiers?
- Azure AI Vision
- Azure AI Custom Vision (Correct)
- Azure AI Face
- Azure AI Video Indexer
Explanation: Azure AI Custom Vision allows developers to upload labeled image data to train customized image classification and object detection models.
Which two capabilities are supported by the Azure AI Face service? (Choose two.)
- Object detection
- Face detection (Correct)
- Speech recognition
- Facial recognition (Correct)
Explanation: Azure AI Face detects human faces, landmarks, and attributes within images, and provides facial verification/identification algorithms.
You need an automated solution to extract structured key-value pairs, tables, and totals from vendor receipts and invoices. Which service should you choose?
- Azure AI Vision OCR
- Azure AI Language
- Azure AI Document Intelligence (Correct)
- Azure AI Search
Explanation: Azure AI Document Intelligence (formerly Form Recognizer) provides prebuilt machine learning models tailored for structured document analysis such as invoices, receipts, and ID documents.
In conversational language understanding (CLU), what term represents the action or goal that a user wants to perform?
- Entity
- Intent (Correct)
- Utterance
- Domain
Explanation: An intent represents the purpose, action, or goal expressed in a user's utterance (e.g., BookFlight or CheckWeather). Entities represent specific parameter values extracted from the utterance.
Which Responsible AI principle states that AI systems should be understandable, clear in purpose, and open about their behavioral limitations?
- Transparency (Correct)
- Accountability
- Privacy and security
- Reliability and safety
Explanation: Transparency ensures AI solutions are accompanied by clear explanations regarding how algorithms work and how data is evaluated.
What feature of Azure Machine Learning automates the process of algorithm selection, feature engineering, and hyperparameter tuning to find the best model?
- Azure Machine Learning designer
- Jupyter Notebooks
- Automated ML (Correct)
- Data labeling
Explanation: Automated ML (AutoML) iterates across machine learning algorithms and hyperparameter combinations to produce the optimal trained model based on your target evaluation metric.
Which computer vision task involves categorizing individual objects and identifying their precise spatial location via bounding boxes?
- Image classification
- Optical character recognition
- Semantic segmentation
- Object detection (Correct)
Explanation: Object detection combines image classification with localization, drawing bounding boxes around individual objects detected in an image.
A travel application needs to translate user comments between multiple languages dynamically. Which dedicated Azure service should be integrated?
- Azure AI Language
- Azure AI Translator (Correct)
- Azure AI Speech
- Azure AI Bot Service
Explanation: Azure AI Translator is a cloud service designed specifically for multi-language text-to-text translation.
Which two capabilities are core features of the Azure AI Speech service? (Choose two.)
- Speech-to-text (Correct)
- Image classification
- Text-to-speech (Correct)
- Object detection
Explanation: Azure AI Speech provides speech-to-text (transcription) and text-to-speech (voice synthesis) APIs.
Which Responsible AI principle states that AI applications should empower everyone and be accessible to people with diverse physical or cognitive capabilities?
- Fairness
- Transparency
- Inclusiveness (Correct)
- Accountability
Explanation: Inclusiveness requires AI technology to be designed to engage and support all individuals, including people with disabilities.
You want to create a conversational knowledge base by extracting question-and-answer pairs directly from an online FAQ URL. Which Azure AI Language feature should you use?
- Conversational language understanding
- Question answering (Correct)
- Key phrase extraction
- Text summarization
Explanation: Question answering in Azure AI Language generates a searchable knowledge base of QA pairs ingested from documents, websites, and manual input.
What data must you provide when training an image classification model using Azure AI Custom Vision?
- Images and tags (Correct)
- Images and bounding box coordinates
- Text and sentiments
- Audio and transcripts
Explanation: Image classification requires a training dataset containing sample images and their respective descriptive tags (labels).
Which metric is standard for evaluating the accuracy and error rate of a regression model?
- Accuracy
- Precision
- Root Mean Squared Error (RMSE) (Correct)
- Silhouette score
Explanation: Root Mean Squared Error (RMSE) evaluates regression models by measuring the difference between predicted values and actual numeric values. Metrics like Accuracy and Precision evaluate classification models.
What is the core principle of Accountability within the Microsoft Responsible AI framework?
- AI systems must be secure from malicious attacks
- Humans are ultimately responsible for the outcomes of an AI system (Correct)
- AI algorithms must be published for open review
- AI systems must never make mistakes
Explanation: Accountability means that the designers, engineers, and organizations that deploy AI systems remain legally and ethically accountable for their systems' decisions and impact.
Which Azure AI Language feature is designed to identify and mask sensitive personal data such as credit card numbers, phone numbers, and addresses?
- PII detection in Azure AI Language (Correct)
- Content Safety
- Sentiment analysis
- Conversational language understanding
Explanation: Personally Identifiable Information (PII) detection identifies and redacts private personal details in unstructured textual data.
Which two resources are managed workspace assets in Azure Machine Learning? (Choose two.)
- Virtual networks
- Compute targets (Correct)
- Data assets (Correct)
- Resource groups
Explanation: Compute targets and Data assets (datastores/datasets) are fundamental resources managed within an Azure Machine Learning workspace.
Which computer vision feature automatically creates a human-readable descriptive sentence summarizing the scene depicted in an image?
- Object detection
- OCR
- Image classification
- Image description (Correct)
Explanation: Image description (captioning) evaluates visual elements in an image and generates a descriptive sentence summarizing the scene.
You submit the sentence: "The camera is exceptional, but the battery life is terrible." to Azure AI Language Sentiment Analysis. What document-level sentiment is returned?
- Positive
- Negative
- Mixed (Correct)
- Neutral
Explanation: Azure AI Language sentiment analysis returns a mixed sentiment score when an input contains distinct positive and negative sentiments across clauses.
Which Responsible AI principle focuses on protecting customer data and preventing unauthorized access or data leaks?
- Privacy and security (Correct)
- Inclusiveness
- Accountability
- Reliability and safety
Explanation: Privacy and security requires that AI data pipelines protect user privacy and adhere to stringent cybersecurity standards.
In Azure Machine Learning, what is the primary role of a compute cluster?
- To store machine learning datasets
- To provide scalable processing power for model training (Correct)
- To visually design machine learning pipelines
- To securely store connection credentials
Explanation: A compute cluster provides auto-scaling multi-node virtual machine infrastructure to run model training jobs and batch processing.
You need to extract spoken dialogue, identify on-screen faces, and extract visual text from video archives. Which Azure AI service provides this unified capability?
- Azure AI Vision
- Azure AI Speech
- Azure Video Indexer (Correct)
- Azure AI Custom Vision
Explanation: Azure Video Indexer uses AI algorithms to analyze and extract metadata, transcripts, visual landmarks, and entities from video and audio files.
Which two data sources can be used to populate a question answering knowledge base in Azure AI Language? (Choose two.)
- Source documents or URLs (Correct)
- A pre-trained speech model
- A set of labeled images
- An optional chit-chat data source (Correct)
Explanation: Question answering knowledge bases can ingest FAQ URLs, files (PDF, DOCX, TSV), and pre-built conversational chit-chat datasets.
You are training a model to predict whether a customer will churn or stay (two possible outcomes). Which machine learning task is this?
- Regression
- Binary classification (Correct)
- Multiclass classification
- Clustering
Explanation: Binary classification classifies data points into exactly one of two mutually exclusive categories.
Which Azure service provides a centralized framework to connect bot applications with external channels like Microsoft Teams, Web Chat, and Slack?
- Azure Logic Apps
- Azure AI Language
- Azure Bot Service (Correct)
- Azure Service Bus
Explanation: Azure Bot Service hosts bot endpoints and handles multi-channel connectivity across communication channels.
In conversational language understanding, what term defines a specific parameter or variable extracted from user input (e.g., "Seattle" in "Book a flight to Seattle")?
- Intent
- Utterance
- Domain
- Entity (Correct)
Explanation: An entity is a specific data point, parameter, or subject identified within an utterance.
Which two services belong to the Azure AI Services product suite? (Choose two.)
- Azure AI Vision (Correct)
- Azure AI Speech (Correct)
- Azure Virtual Machines
- Azure SQL Database
Explanation: Azure AI Vision and Azure AI Speech are prebuilt cloud AI services within Azure AI Services. Azure Virtual Machines and Azure SQL Database are infrastructure and relational data services.
Which Azure Machine Learning capability records run metrics, hyperparameters, and generated model artifacts during training?
- Experiments (Correct)
- Endpoints
- Compute instances
- Data labeling
Explanation: Experiments in Azure Machine Learning organize and record logs, parameters, metrics, and output files from individual training runs.
Which Azure AI Speech capability synthesizes text strings into spoken audio output?
- Speech-to-text
- Translation
- Text-to-speech (Correct)
- Key phrase extraction
Explanation: Text-to-speech (TTS) converts text into natural synthetic speech across various voices and languages.
You need to segment a retail customer base into behavior patterns where no existing category labels exist. What machine learning paradigm should you implement?
- Regression
- Clustering (Correct)
- Classification
- Supervised learning
Explanation: Clustering is an unsupervised learning technique that groups unlabeled records into distinct clusters based on mathematical similarity.
Which file format cannot be directly uploaded as a structured FAQ knowledge base in Azure AI Language Question Answering?
- PDF
- DOCX
- JPG (Correct)
- TSV
Explanation: Question Answering directly accepts text documents like PDF, DOCX, TSV, and TXT, but does not support image formats like JPG for question extraction.
Which three visual capabilities are supported by Azure AI Vision? (Choose three.)
- Optical Character Recognition (OCR) (Correct)
- Object detection (Correct)
- Spatial analysis (Correct)
- Sentiment analysis
Explanation: Azure AI Vision provides OCR, object detection, and spatial analysis. Sentiment analysis is a natural language service.
Which machine learning approach trains algorithms using historical data that already contains known output labels?
- Unsupervised learning
- Supervised learning (Correct)
- Reinforcement learning
- Zero-shot learning
Explanation: Supervised machine learning trains algorithms on labeled data where known outputs guide pattern recognition.
You need to deploy a trained machine learning model so client web applications can invoke it for immediate, low-latency inferencing. What should you create in Azure Machine Learning?
- Real-time endpoint (Correct)
- Batch endpoint
- Compute cluster
- Data asset
Explanation: Real-time managed endpoints host models as online REST APIs for immediate inferencing.
Which Azure AI Language feature evaluates unstructured text to identify main talking points and key conceptual phrases?
- Language detection
- Key phrase extraction (Correct)
- Sentiment analysis
- Translation
Explanation: Key phrase extraction identifies and returns the principal talking points and terms within an unstructured body of text.
An automated credit approval model denies loans to one demographic group significantly more often than another with identical income and credit profiles. Which Responsible AI principle is violated?
- Reliability and safety
- Transparency
- Accountability
- Fairness (Correct)
Explanation: The Fairness principle requires that AI systems treat individuals without algorithmic bias or disparate demographic discrimination.
Which integrated tool in Azure Machine Learning studio provides an interactive Jupyter environment for exploratory data science with Python?
- Notebooks (Correct)
- Designer
- Automated ML
- Endpoints
Explanation: Azure Machine Learning Notebooks host managed Jupyter environments attached directly to compute instances.
To make API calls to an Azure AI Service resource, what information must client applications supply alongside the service endpoint URL?
- Virtual network ID
- Resource group name
- Subscription key (Correct)
- Azure Active Directory Object ID
Explanation: Azure AI Services REST APIs authenticate requests using an authorized subscription key or Azure Entra token passed in the request headers.
In Azure Machine Learning designer, what connects data preparation, training, and scoring modules together in a defined execution workflow?
- Compute instance
- Pipeline (Correct)
- Datastore
- Endpoint
Explanation: A pipeline in Azure Machine Learning designer visually orchestrates data flow and computation steps between connected modules.
Which two data exchange formats are supported for payload transmission in Azure AI Services REST APIs? (Choose two.)
- CSV
- JSON (Correct)
- XML (Correct)
- YAML
Explanation: Azure AI Services REST endpoints primarily accept and return standard structured JSON payloads, with certain services (like speech synthesis/SSML) accepting XML payloads.
Which Azure AI service provides prebuilt models specifically trained to analyze and extract information from receipts, business cards, and identity documents?
- Azure AI Document Intelligence (Correct)
- Azure AI Vision
- Azure AI Search
- Azure AI Custom Vision
Explanation: Azure AI Document Intelligence provides pre-trained models tailored specifically for extracting structured key fields from standard business documents like receipts, business cards, and ID cards.