dc dotCreds
Microsoft Azure AI Fundamentals (AI-901)

AI-901 Practice Test

Start a free 30-question AI-901 daily set with source-backed explanations, local progress, and a fresh rotation every morning.

30 daily web questions Source-backed explanations 7-day score history Questions updated at Apr 14, 2026, 12:35 PM CDT
AI-901 icon

AI-901

Microsoft Azure AI Fundamentals (AI-901)

Why this page works

  • Thirty focused questions every day
  • Source links on every explanation
  • Local progress saved automatically
  • Email sync path ready for later
  • Apps provide deeper drills when available
Today's 30 AI-901 questions

Use this AI-901 practice test to review Microsoft AI Fundamentals. Questions rotate daily and each answer links back to the source used to write it.

Today’s Set
30 questions
Daily set rotates at 10:00 AM local time
Progress
0/30
Answered on this page session
Accuracy
0%
Loading countdown…

7-day score keeper

Answer questions today and this will become a rolling 7-day scorecard.

Local history
Optional progress sync

Keep today’s practice moving

Guest progress saves automatically on this device. Add an email later when you want a magic link that keeps your daily AI-901 practice in sync across browsers.

Guest progress saves on this device automatically

Guest progress is available without an account.

100 verified questions are currently in the live bank. Questions updated at Apr 14, 2026, 12:35 PM CDT. The daily set rotates at 10:00 AM local time, and each explanation links back to the source used to write it. Use the web set for quick practice, then switch to the app when available for larger banks and deeper review.

Official exam resources

Use these official Microsoft resources alongside the daily practice set. They cover the provider's own exam page, study guide, or prep material.

Need adjacent Microsoft practice pages too? Microsoft practice hub.

Question 1 of 30
Objective 3.2 Describe computer vision and language workloads on Azure

Which Azure AI Language feature identifies people, places, and organizations mentioned in text?

Concept tested: Describe computer vision and language workloads on Azure

A. Incorrect: Sentiment analysis is incorrect because it does not answer this stem as directly as Named entity recognition.

B. Correct: Named entity recognition is the correct answer because named entity recognition identifies entities such as people, locations, and organizations in text.

C. Incorrect: Language detection is incorrect because it does not answer this stem as directly as Named entity recognition.

D. Incorrect: Image analysis is incorrect because it does not answer this stem as directly as Named entity recognition.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 2 of 30
Objective 1.1 Describe AI workloads and considerations

A developer needs to convert live meeting audio into text and also support spoken translation. Which Azure AI service is the best fit?

Concept tested: Describe AI workloads and considerations

A. Incorrect: Azure AI Search is incorrect because it does not answer this stem as directly as Azure Speech.

B. Correct: Azure Speech is the correct answer because azure Speech supports speech to text, text to speech, and speech translation scenarios, so it is the right service for transcribing and translating spoken audio.

C. Incorrect: Azure Document Intelligence is incorrect because it does not answer this stem as directly as Azure Speech.

D. Incorrect: Azure AI Content Safety is incorrect because it does not answer this stem as directly as Azure Speech.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 3 of 30
Objective 4.2 Describe generative AI fundamentals

In a RAG application, what is the role of an index?

Concept tested: Describe generative AI fundamentals

A. Correct: Organize content so relevant passages can be retrieved efficiently is the correct answer because an index organizes searchable content so the app can efficiently retrieve relevant grounding data for a model response.

B. Incorrect: Generate user accounts for every prompt is incorrect because it does not answer this stem as directly as Organize content so relevant passages can be retrieved efficiently..

C. Incorrect: Disable semantic search and citations is incorrect because it does not answer this stem as directly as Organize content so relevant passages can be retrieved efficiently..

D. Incorrect: Replace grounding data with model training data is incorrect because it does not answer this stem as directly as Organize content so relevant passages can be retrieved efficiently..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 4 of 30
Objective 2.1 Describe machine learning fundamentals

What is the main benefit of automated machine learning in Azure Machine Learning?

Concept tested: Describe machine learning fundamentals

A. Correct: It automates iterative model-development tasks such as trying algorithms and parameters is the correct answer because automated ML automates time-consuming model-development tasks by trying different algorithms, parameters, and feature selections during training.

B. Incorrect: It removes the need for labeled training data in every machine learning project is incorrect because it does not answer this stem as directly as It automates iterative model-development tasks such as trying algorithms and parameters..

C. Incorrect: It guarantees that every model is ready for production without evaluation is incorrect because it does not answer this stem as directly as It automates iterative model-development tasks such as trying algorithms and parameters..

D. Incorrect: It converts all AI workloads into relational database workloads is incorrect because it does not answer this stem as directly as It automates iterative model-development tasks such as trying algorithms and parameters..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 5 of 30
Objective 5.1 Describe AI governance, security, and operations

Which content-filtering risk is related to malicious instructions hidden inside retrieved documents used for grounding?

Concept tested: Describe AI governance, security, and operations

A. Correct: Indirect prompt attack is the correct answer because indirect attacks can occur when malicious instructions are embedded in documents or retrieved content that a generative AI system processes as grounding data.

B. Incorrect: Image smart crop is incorrect because it does not answer this stem as directly as Indirect prompt attack.

C. Incorrect: Regression error is incorrect because it does not answer this stem as directly as Indirect prompt attack.

D. Incorrect: Batch transcription is incorrect because it does not answer this stem as directly as Indirect prompt attack.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 6 of 30
Objective 3.3 Describe computer vision and language workloads on Azure

Which Azure Speech capability generates synthetic spoken audio from text input?

Concept tested: Describe computer vision and language workloads on Azure

A. Correct: Text to speech is the correct answer because text to speech generates synthesized spoken audio from text input using Azure Speech Service.

B. Incorrect: Speech to text is incorrect because it does not answer this stem as directly as Text to speech.

C. Incorrect: Speech translation is incorrect because it does not answer this stem as directly as Text to speech.

D. Incorrect: Speaker diarization is incorrect because it does not answer this stem as directly as Text to speech.

Why this matters: This matters because Describe computer vision and language workloads on Azure questions test whether Text to speech fits the scenario's constraints, not just whether the term sounds familiar.
Question 7 of 30
Objective 1.3 Describe AI workloads and considerations

Which scenario is best suited for a rule-based system?

Concept tested: Describe AI workloads and considerations

A. Incorrect: Analyzing customer reviews to detect sentiment is incorrect because it does not answer this stem as directly as Determining if an email is spam or not spam.

B. Incorrect: Predicting stock prices based on historical data is incorrect because it does not answer this stem as directly as Determining if an email is spam or not spam.

C. Correct: Determining if an email is spam or not spam is the correct answer because rule-based systems are ideal for tasks that require deterministic logic and specific rules, such as filtering spam emails.

D. Incorrect: Identifying patterns in medical images is incorrect because it does not answer this stem as directly as Determining if an email is spam or not spam.

Why this matters: This matters because Microsoft 365 admin questions test which admin area changes mail, groups, users, or collaboration settings.
Question 8 of 30
Objective 4.3 Describe generative AI fundamentals

In a retrieval-augmented generation system, what does vector search do directly?

Concept tested: Describe generative AI fundamentals

A. Incorrect: To store and manage data efficiently is incorrect because it does not answer this stem as directly as To retrieve relevant information quickly.

B. Correct: To retrieve relevant information quickly is the correct answer because vector search directly retrieves the most relevant content from an index based on similarity. That retrieved context can then improve the quality of the generated answer.

C. Incorrect: To enhance the quality of generated content is incorrect because it does not answer this stem as directly as To retrieve relevant information quickly.

D. Incorrect: To ensure data privacy is incorrect because it does not answer this stem as directly as To retrieve relevant information quickly.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 9 of 30
Objective 2.3 Describe machine learning fundamentals

Which Azure service is specifically designed to manage and automate the lifecycle of machine learning models?

Concept tested: Describe machine learning fundamentals

A. Incorrect: Azure Kubernetes Service (AKS) is incorrect because it does not answer this stem as directly as Azure Machine Learning.

B. Incorrect: Azure Functions is incorrect because it does not answer this stem as directly as Azure Machine Learning.

C. Correct: Azure Machine Learning is the correct answer because azure Machine Learning provides a platform for building, training, and deploying machine learning models.

D. Incorrect: Azure Logic Apps is incorrect because it does not answer this stem as directly as Azure Machine Learning.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 10 of 30
Objective 5.2 Describe AI governance, security, and operations

What does AI observability help teams do across the AI application lifecycle?

Concept tested: Describe AI governance, security, and operations

A. Correct: Monitor, evaluate, trace, and troubleshoot AI systems is the correct answer because aI observability helps teams monitor, understand, evaluate, trace, and troubleshoot AI systems across development and production.

B. Incorrect: Skip pre-production testing and deploy faster is incorrect because it does not answer this stem as directly as Monitor, evaluate, trace, and troubleshoot AI systems..

C. Incorrect: Remove the need for quality or safety metrics is incorrect because it does not answer this stem as directly as Monitor, evaluate, trace, and troubleshoot AI systems..

D. Incorrect: Use only screenshots instead of logs or traces is incorrect because it does not answer this stem as directly as Monitor, evaluate, trace, and troubleshoot AI systems..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 11 of 30
Objective 3.1 Describe computer vision and language workloads on Azure

Which Azure AI service can return captions, tags, and detected objects from an image without training a custom model?

Concept tested: Describe computer vision and language workloads on Azure

A. Correct: Azure AI Vision is the correct answer because azure AI Vision image analysis can describe images and extract visual features such as captions, tags, and detected objects without custom training.

B. Incorrect: Azure AI Language is incorrect because it does not answer this stem as directly as Azure AI Vision.

C. Incorrect: Azure Machine Learning is incorrect because it does not answer this stem as directly as Azure AI Vision.

D. Incorrect: Azure Cosmos DB is incorrect because it does not answer this stem as directly as Azure AI Vision.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 12 of 30
Objective 1.2 Describe AI workloads and considerations

A healthcare provider is implementing a chatbot to assist patients with scheduling appointments. What should they consider to ensure the reliability of their AI system?

Concept tested: Describe AI workloads and considerations

A. Incorrect: Allow the chatbot to make decisions about patient diagnoses without human oversight is incorrect because it does not answer this stem as directly as Regularly update the chatbot's training data to include the latest medical information..

B. Correct: Regularly update the chatbot's training data to include the latest medical information is the correct answer because regularly update the chatbot's training data to include the latest medical information. Reliability requires the AI system to provide accurate and consistent information. Regular updates ensure this.

C. Incorrect: Use a simple, one-size-fits-all response template for all queries is incorrect because it does not answer this stem as directly as Regularly update the chatbot's training data to include the latest medical information..

D. Incorrect: Ignore potential errors in the chatbot's responses as they are minor is incorrect because it does not answer this stem as directly as Regularly update the chatbot's training data to include the latest medical information..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 13 of 30
Objective 4.1 Describe generative AI fundamentals

What is the primary purpose of a context window in a language model?

Concept tested: Describe generative AI fundamentals

A. Incorrect: To store historical data for analysis is incorrect because it does not answer this stem as directly as To provide additional information to improve the quality of generated text..

B. Incorrect: To limit the amount of text the model can generate at once is incorrect because it does not answer this stem as directly as To provide additional information to improve the quality of generated text..

C. Correct: To provide additional information to improve the quality of generated text is the correct answer because to provide additional information to improve the quality of generated text. A context window in a language model provides additional information to help it understand and generate more accurate and relevant text based on the input provided.

D. Incorrect: To define the geographical location where the model operates is incorrect because it does not answer this stem as directly as To provide additional information to improve the quality of generated text..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 14 of 30
Objective 2.2 Describe machine learning fundamentals

Which metric is commonly used to evaluate the precision of a classification model?

Concept tested: Describe machine learning fundamentals

A. Incorrect: Accuracy is incorrect because it does not answer this stem as directly as Precision.

B. Incorrect: Recall is incorrect because it does not answer this stem as directly as Precision.

C. Correct: Precision is the correct answer because precision measures the proportion of true positive predictions out of all positive predictions made by the model.

D. Incorrect: F1 Score is incorrect because it does not answer this stem as directly as Precision.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 15 of 30
Objective 5.3 Describe AI governance, security, and operations

Which of the following best describes a policy control in an AI project?

Concept tested: Describe AI governance, security, and operations

A. Incorrect: A document outlining the technical architecture is incorrect because it does not answer this stem as directly as A set of guidelines for ethical use and deployment..

B. Incorrect: A tool for encrypting data at rest is incorrect because it does not answer this stem as directly as A set of guidelines for ethical use and deployment..

C. Correct: A set of guidelines for ethical use and deployment is the correct answer because a set of guidelines for ethical use and deployment. A policy control is a set of guidelines for ethical use and deployment, which helps ensure that AI projects are conducted responsibly.

D. Incorrect: A script for automating model training is incorrect because it does not answer this stem as directly as A set of guidelines for ethical use and deployment..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 16 of 30
Objective 3.1 Describe computer vision and language workloads on Azure

Which Azure AI Vision capability extracts printed and handwritten text from images?

Concept tested: Describe computer vision and language workloads on Azure

A. Incorrect: Object detection is incorrect because it does not answer this stem as directly as OCR.

B. Correct: OCR is the correct answer because oCR in Azure AI Vision reads printed and handwritten text from images and documents.

C. Incorrect: Image captions is incorrect because it does not answer this stem as directly as OCR.

D. Incorrect: Face detection is incorrect because it does not answer this stem as directly as OCR.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 17 of 30
Objective 1.1 Describe AI workloads and considerations

A company needs to analyze customer reviews to identify common themes and sentiment. Which Azure AI service is most appropriate for this task?

Concept tested: Describe AI workloads and considerations

A. Correct: Azure AI Language is the correct answer because azure AI Language provides text analysis capabilities such as sentiment analysis and key phrase extraction for unstructured text.

B. Incorrect: Azure Machine Learning is incorrect because it does not answer this stem as directly as Azure AI Language.

C. Incorrect: Azure Functions is incorrect because it does not answer this stem as directly as Azure AI Language.

D. Incorrect: Azure Logic Apps is incorrect because it does not answer this stem as directly as Azure AI Language.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 18 of 30
Objective 4.3 Describe generative AI fundamentals

In prompt construction, what is the main purpose of instructions?

Concept tested: Describe generative AI fundamentals

A. Correct: Tell the model what task to perform and shape the expected response is the correct answer because instructions are a core prompt component that tells the model what to do and helps shape the response toward the desired task.

B. Incorrect: Store vector embeddings in a search index is incorrect because it does not answer this stem as directly as Tell the model what task to perform and shape the expected response..

C. Incorrect: Train a new speech recognition model is incorrect because it does not answer this stem as directly as Tell the model what task to perform and shape the expected response..

D. Incorrect: Assign cloud resource permissions is incorrect because it does not answer this stem as directly as Tell the model what task to perform and shape the expected response..

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 19 of 30
Objective 2.3 Describe machine learning fundamentals

What is the primary purpose of a dataset in Azure Machine Learning?

Concept tested: Describe machine learning fundamentals

A. Incorrect: To store and manage machine learning models is incorrect because it does not answer this stem as directly as To provide input data for model training and testing.

B. Correct: To provide input data for model training and testing is the correct answer because datasets in Azure Machine Learning are used to provide input data for training and testing machine learning models.

C. Incorrect: To define the training parameters for algorithms is incorrect because it does not answer this stem as directly as To provide input data for model training and testing.

D. Incorrect: To monitor the performance of deployed models is incorrect because it does not answer this stem as directly as To provide input data for model training and testing.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 20 of 30
Objective 5.1 Describe AI governance, security, and operations

Which four harm categories are covered by the neural classification models in Microsoft Foundry content filtering?

Concept tested: Describe AI governance, security, and operations

A. Correct: Hate, sexual, violence, and self-harm is the correct answer because microsoft Foundry content filtering includes neural classification models for hate, sexual, violence, and self-harm content across severity levels.

B. Incorrect: Billing, storage, compute, and networking is incorrect because it does not answer this stem as directly as Hate, sexual, violence, and self-harm.

C. Incorrect: Forecasting, clustering, regression, and ranking is incorrect because it does not answer this stem as directly as Hate, sexual, violence, and self-harm.

D. Incorrect: Images, tables, charts, and forms is incorrect because it does not answer this stem as directly as Hate, sexual, violence, and self-harm.

Why this matters: This matters because agent-administration questions test whether hosting, orchestration, and workflow terms match how Copilot agents are deployed.
Question 21 of 30
Objective 3.2 Describe computer vision and language workloads on Azure

Which Azure Language feature returns the main concepts from a document as a list?

Concept tested: Describe computer vision and language workloads on Azure

A. Correct: Key phrase extraction is the correct answer because key phrase extraction evaluates text and returns the main concepts or important phrases from unstructured content.

B. Incorrect: Image captioning is incorrect because it does not answer this stem as directly as Key phrase extraction.

C. Incorrect: Speech translation is incorrect because it does not answer this stem as directly as Key phrase extraction.

D. Incorrect: Object detection is incorrect because it does not answer this stem as directly as Key phrase extraction.

Why this matters: This matters because Describe computer vision and language workloads on Azure questions test whether Key phrase extraction fits the scenario's constraints, not just whether the term sounds familiar.
Question 22 of 30
Objective 1.1 Describe AI workloads and considerations

Which Foundry Tool should a team use to detect harmful text and image content in user-generated prompts and AI-generated responses?

Concept tested: Describe AI workloads and considerations

A. Correct: Azure AI Content Safety is the correct answer because azure AI Content Safety provides APIs for detecting harmful text and image content, including content submitted by users and generated by AI systems.

B. Incorrect: Azure AI Search is incorrect because it does not answer this stem as directly as Azure AI Content Safety.

C. Incorrect: Azure Machine Learning AutoML is incorrect because it does not answer this stem as directly as Azure AI Content Safety.

D. Incorrect: Azure Document Intelligence is incorrect because it does not answer this stem as directly as Azure AI Content Safety.

Why this matters: This matters because agent-administration questions test whether hosting, orchestration, and workflow terms match how Copilot agents are deployed.
Question 23 of 30
Objective 4.2 Describe generative AI fundamentals

What is the purpose of embeddings in Azure OpenAI services?

Concept tested: Describe generative AI fundamentals

A. Incorrect: To filter inappropriate content is incorrect because it does not answer this stem as directly as To represent data as numerical vectors.

B. Incorrect: To generate human-like text is incorrect because it does not answer this stem as directly as To represent data as numerical vectors.

C. Incorrect: To limit API usage is incorrect because it does not answer this stem as directly as To represent data as numerical vectors.

D. Correct: To represent data as numerical vectors is the correct answer because embeddings are used to represent data as numerical vectors, which can be useful for various AI applications.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 24 of 30
Objective 2.1 Describe machine learning fundamentals

A model predicts whether a support ticket is billing, technical, or account-related. What type of supervised learning task is this?

Concept tested: Describe machine learning fundamentals

A. Correct: Classification is the correct answer because classification predicts which category a new item belongs to, such as assigning a support ticket to one of several ticket classes.

B. Incorrect: Regression is incorrect because it does not answer this stem as directly as Classification.

C. Incorrect: Clustering is incorrect because it does not answer this stem as directly as Classification.

D. Incorrect: Anomaly detection is incorrect because it does not answer this stem as directly as Classification.

Why this matters: This matters because Copilot administration questions separate license assignment, billing policy, analytics, and security controls.
Question 25 of 30
Objective 5.2 Describe AI governance, security, and operations

A developer is working on an AI model that processes sensitive customer data. What is the most important step to ensure the model's safety?

Concept tested: Describe AI governance, security, and operations

A. Incorrect: Regularly update the model architecture is incorrect because it does not answer this stem as directly as Implement robust data encryption and access controls.

B. Correct: Implement robust data encryption and access controls is the correct answer because data encryption and access controls are crucial for protecting sensitive information.

C. Incorrect: Conduct thorough quality metrics analysis is incorrect because it does not answer this stem as directly as Implement robust data encryption and access controls.

D. Incorrect: Perform A/B testing with different datasets is incorrect because it does not answer this stem as directly as Implement robust data encryption and access controls.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 26 of 30
Objective 3.1 Describe computer vision and language workloads on Azure

Which Azure AI Vision capability identifies and locates multiple objects within an image?

Concept tested: Describe computer vision and language workloads on Azure

A. Incorrect: Image captions is incorrect because it does not answer this stem as directly as Object detection.

B. Incorrect: Optical character recognition is incorrect because it does not answer this stem as directly as Object detection.

C. Correct: Object detection is the correct answer because object detection in Azure AI Vision identifies objects and returns their locations in the image.

D. Incorrect: Sentiment analysis is incorrect because it does not answer this stem as directly as Object detection.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 27 of 30
Objective 1.1 Describe AI workloads and considerations

Which Azure service is best suited for real-time anomaly detection in sensor data?

Concept tested: Describe AI workloads and considerations

A. Correct: Azure Stream Analytics is the correct answer because azure Stream Analytics is designed for real-time processing and analysis of large volumes of data, making it ideal for anomaly detection in sensor data.

B. Incorrect: Azure Cognitive Services is incorrect because it does not answer this stem as directly as Azure Stream Analytics.

C. Incorrect: Azure Machine Learning is incorrect because it does not answer this stem as directly as Azure Stream Analytics.

D. Incorrect: Azure Functions is incorrect because it does not answer this stem as directly as Azure Stream Analytics.

Why this matters: This matters because Copilot administration questions separate license assignment, billing policy, analytics, and security controls.
Question 28 of 30
Objective 4.1 Describe generative AI fundamentals

Which of the following is an example of prompt engineering?

Concept tested: Describe generative AI fundamentals

A. Incorrect: Designing a user interface for a chatbot is incorrect because it does not answer this stem as directly as Creating a set of instructions to guide the generation of text by a language model..

B. Incorrect: Writing code to integrate AI services into an application is incorrect because it does not answer this stem as directly as Creating a set of instructions to guide the generation of text by a language model..

C. Incorrect: Developing algorithms to improve image recognition accuracy is incorrect because it does not answer this stem as directly as Creating a set of instructions to guide the generation of text by a language model..

D. Correct: Creating a set of instructions to guide the generation of text by a language model is the correct answer because creating a set of instructions to guide the generation of text by a language model. Prompt engineering involves creating specific and clear instructions or prompts to guide a language model in generating the desired output, improving the quality and relevance of the generated text.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 29 of 30
Objective 2.2 Describe machine learning fundamentals

What is the F1 Score and how does it combine precision and recall?

Concept tested: Describe machine learning fundamentals

A. Incorrect: The F1 Score is the average of precision and recall is incorrect because it does not answer this stem as directly as The F1 Score is the harmonic mean of precision and recall..

B. Incorrect: The F1 Score is the product of precision and recall is incorrect because it does not answer this stem as directly as The F1 Score is the harmonic mean of precision and recall..

C. Incorrect: The F1 Score is a weighted sum of precision and recall is incorrect because it does not answer this stem as directly as The F1 Score is the harmonic mean of precision and recall..

D. Correct: The F1 Score is the harmonic mean of precision and recall is the correct answer because the F1 Score is the harmonic mean of precision and recall. The F1 Score is the harmonic mean of precision and recall, providing a single metric that balances both measures.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Question 30 of 30
Objective 5.1 Describe AI governance, security, and operations

What is the purpose of network isolation in Azure?

Concept tested: Describe AI governance, security, and operations

A. Incorrect: To increase performance is incorrect because it does not answer this stem as directly as To enhance security by limiting access to resources.

B. Correct: To enhance security by limiting access to resources is the correct answer because network isolation in Azure helps enhance security by limiting access to resources and preventing unauthorized access.

C. Incorrect: To reduce costs is incorrect because it does not answer this stem as directly as To enhance security by limiting access to resources.

D. Incorrect: To simplify management is incorrect because it does not answer this stem as directly as To enhance security by limiting access to resources.

Why this matters: This matters because AI questions test whether the control changes model behavior, data handling, or evaluation in the way the scenario requires.
Where to go after the daily web set

How are AI-901 questions generated?

dotCreds builds AI-901 practice questions from public exam objectives and Microsoft Learn and exam-objective references. The questions are written for realistic study practice, not copied from exam dumps.

How are explanations sourced?

Each question includes an explanation and, when available, a source link back to the provider documentation or reference used to validate the answer. That keeps the practice tied to study material you can actually review.

What score do I get?

The page tracks today's answered count and accuracy for the 30-question daily set, then saves a 7-day score history on this device so you can see your recent practice trend.

Why use this site?

The site is the fastest way to start AI-901 practice without installing anything. It is built for daily recall, quick weak-topic discovery, and source-backed explanations you can review immediately.

Why use the app when available?

The web page is the quick free sampler. If a dotCreds app is available for AI-901, the app is better for larger banks, focused weak-domain drills, longer review sessions, and mobile study routines.