Microsoft AI-900 Online Practice Questions and Exam Preparation

Microsoft AI-900 Online Questions & Answers

• Question 331:

  HOTSPOT

  Select the answer that correctly completes the sentence.

  

  

  ---

  In Microsoft's Responsible AI framework, the Reliability and Safety principle ensures that AI systems perform consistently, safely, and as intended across diverse conditions - even when faced with incomplete, unusual, or unexpected data.

  Correctly handling unusual or missing values in a dataset directly demonstrates this principle, as it helps prevent faulty predictions, biased results, or unsafe system behaviors. According to the Microsoft Learn Responsible AI module (from the AI-900 and AI-102 study paths), a reliable AI model should maintain its performance when encountering data anomalies. This includes validating inputs, managing missing or extreme values, and testing models to ensure they behave as expected in real-world scenarios. Such practices make AI systems robust and trustworthy, which aligns exactly with the Reliability and Safety principle. The other Responsible AI principles address different concerns:

  Inclusiveness: Ensures AI empowers and serves all users equitably.

  Privacy and Security: Focuses on safeguarding personal data and preventing unauthorized access.

  Transparency: Ensures that AI decisions are understandable and explainable to users. While all principles are essential, managing data integrity and system stability-including how a model responds to missing or anomalous values-is primarily a matter of reliability and safety. It ensures the AI behaves predictably and minimizes risks of errors or unintended harm. Therefore, the correct completion of the sentence is: "Correctly handling unusual or missing values is an example of the application of the Reliability and Safety principle for Responsible AI."

• Question 332:

  You have a dataset that contains experimental data for fuel samples.

  You need to predict the amount of energy that can be obtained from a sample based on its density.

  Which type of Al workload should you use?

  A. Classification
  B. Clustering
  C. Knowledge mining
  D. Regression
  D. Regression

• Question 333:

  Which two components can you drag onto a canvas in Azure Machine Learning designer? Each correct answer presents a complete solution.

  NOTE: Each correct selection is worth one point.

  A. dataset
  B. compute
  C. pipeline
  D. module
  A. dataset
  D. module

  ---

  Explanation

  You can drag-and-drop datasets and modules onto the canvas.

  References:

  https://docs.microsoft.com/en-us/azure/machine-learning/concept-designer

• Question 334:

  HOTSPOT

  For each of the following statements, select Yes if the statement is true. Otherwise, select No.

  NOTE: Each correct selection is worth one point

  

  

• Question 335:

  HOTSPOT

  Select the answer that correctly completes the sentence.

  

  

  ---

  Numeric

  For regression problems, the label column must contain numeric data that represents the response variable. Ideally the numeric data represents a continuous scale.

  According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and the Microsoft Learn module "Identify features of common machine learning types", a regression model is a type of supervised machine learning model that is used to predict continuous numeric values based on one or more input variables (features). In supervised learning, models are trained using labeled data, where each input record has a known target value (label). For regression specifically, the label represents a numeric quantity - such as price, age, temperature, or sales figures. The model learns to approximate a function that maps input variables to numeric outputs. For example: Predicting house prices based on size, location, and number of rooms. Predicting monthly sales revenue from marketing spend and seasonality. Forecasting temperature based on historical weather data. In Azure Machine Learning Designer and AutoML, when building a regression model, the label column must therefore be of numeric data type (integer or float). If the label were categorical (for example, "yes/no" or "approved/denied"), the appropriate model type would be classification, not regression. To contrast with other options: Boolean - used in binary classification problems (true/false outcomes). Datetime - used for time series forecasting, not standard regression labels. Text - used as input features for NLP models, not as regression targets. Hence, when configuring a regression task in Azure, ensuring the target variable (label) is numeric is a fundamental requirement. The model's performance metrics-such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-also rely on numeric computations.

  When building a K-means clustering model, all features (variables) used in the model must be numeric in nature. According to the Microsoft Azure AI Fundamentals (AI-900) study materials and standard machine learning theory, K-means clustering is an unsupervised learning algorithm that groups data points into clusters based on their similarity - specifically by minimizing the Euclidean distance between data points and their assigned cluster centroids. Because the K-means algorithm depends on distance calculations, it requires numeric data types. The Euclidean distance (or similar measures) can only be computed between numerical values. Therefore, all categorical or text data must first be converted into numeric form through feature engineering techniques such as one-hot encoding, label encoding, or embedding vectors, depending on the nature of the data. Here's how K-means works in summary: The algorithm initializes a predefined number of centroids (K). Each data point is assigned to the nearest centroid based on numeric distance. The centroids are recalculated as the mean of the points in each cluster. The process repeats until convergence. If non-numeric data (e.g., text or Boolean) were provided, the model would not be able to calculate distances accurately, leading to computational errors. Other options are incorrect: Boolean and integer types can represent numeric values but are considered special cases; the algorithm requires general numeric representation (e.g., continuous values). Text cannot be processed directly without conversion. Thus, according to Azure Machine Learning and AI-900 official concepts, all features in a K- means clustering model must be numeric to ensure valid mathematical operations and clustering accuracy.

  References:

  https://learn.microsoft.com/en-us/azure/machine-learning/component-reference/train-model

• Question 336:

  DRAG DROP

  Match the Al solution to the appropriate task.

  To answer, drag the appropriate solution from the column on the left to its task on the right. Each solution may be used once, more than once, or not at all.

  NOTE: Each correct match is worth one point.

  Select and Place:

  

  

• Question 337:

  You have a webchat bot that provides responses from a QnA Maker knowledge base.

  You need to ensure that the bot uses user feedback to improve the relevance of the responses over time.

  What should you use?

  A. key phrase extraction
  B. sentiment analysis
  C. business logic
  D. active learning
  D. active learning

  ---

  Explanation

  References:

  https://docs.microsoft.com/en-us/azure/cognitive-services/qnamaker/how-to/improve-knowledge-base

• Question 338:

  For which two workloads can you use computer vision?

  Each correct answer presents a complete solution. NOTE: Each correct selection is worth one point.

  A. creating photorealistic images by using three-dimensional models
  B. assigning the color pixels in an image to object names
  C. describing the contents of an image
  D. detecting inconsistencies and anomalies in a stream of data
  E. creating visual representations of numerical data
  B. assigning the color pixels in an image to object names
  C. describing the contents of an image

  ---

  Explanation

  Azure Cognitive Service for Vision is a unified service that offers innovative computer vision capabilities. Give your apps the ability to analyze images, read text, and detect faces with prebuilt image tagging, text extraction with optical character recognition (OCR), and responsible facial recognition. Incorporate vision features into your projects with no machine learning experience required. Azure's Computer Vision service gives you access to advanced algorithms that process images and return information based on the visual features you're interested in.

  * Image Analysis

  The Image Analysis service extracts many visual features from images, such as objects, faces, adult content, and auto-generated text descriptions.

  * OCR

  * Face

  * Spatial Analysis References:

  https://learn.microsoft.com/en-us/azure/cognitive-services/computer-vision/overview

• Question 339:

  You need to provide customers with the ability to query the status of orders by using phones, social media, or digital assistants.

  What should you use?

  A. an Azure Machine Learning model
  B. the Translator service
  C. a Form Recognizer model
  D. Azure Bot Service
  D. Azure Bot Service

  ---

  Explanation

  Azure Bot Service provides an integrated development environment for bot building. Its integration with Power Virtual Agents, a fully hosted low-code platform, enables developers of all technical abilities build conversational AI bots - no code needed.

  References:

  https://azure.microsoft.com/en-us/products/bot-services

• Question 340:

  A developer wants to experiment with prompts for a generative AI model before deploying the solution.

  Where can prompts be evaluated in Azure AI Foundry?

  A. Playground
  B. Azure Monitor
  C. Dataset manager
  D. Resource group
  A. Playground

  ---

  Explanation

  Playground provides an interactive interface where developers can test prompts, adjust parameters, and review responses from generative AI models. Azure Monitor focuses on monitoring resources and performance, not prompt experimentation.