AI-900 Practice Test 12 — 15 Questions

Question 1

What is 'video action recognition' in computer vision?

Accepted Answer

Identifying human activities (running, cooking, falling) from temporal patterns across video frames. Video action recognition is a computer vision technique that analyzes sequences of video frames to identify and classify human activities based on temporal patterns and motion cues. Option B correctly describes this as identifying activities like running, cooking, or falling from temporal patterns across frames, which is the core definition used in Azure Video Indexer and other AI services.

Answer

Recognising which video format (MP4, MOV) an uploaded file uses

Answer

Detecting when inappropriate actions are performed in user-generated video content

Answer

Tracking when viewers take actions (like, share, comment) in response to a video

Question 2

What is the primary use case for Azure AI Vision's 'image retrieval' using multimodal embeddings?

Accepted Answer

Enabling natural language image search and finding visually similar images using semantic understanding. Azure AI Vision's image retrieval using multimodal embeddings is designed to enable natural language image search and find visually similar images by leveraging semantic understanding. It converts both images and text into vector embeddings in a shared semantic space, allowing queries like 'a red car on a beach' to retrieve relevant images without relying on exact keyword matches or pre-defined tags.

Answer

Storing images in Azure Blob Storage with automatic tagging

Answer

Automatically resizing images for different screen sizes

Answer

Detecting copyrighted images in user-uploaded content

Question 3

What is 'data preprocessing' and why is it important for machine learning?

Accepted Answer

Transforming raw data (handling nulls, scaling, encoding) to make it suitable for ML training. Data preprocessing is the transformation of raw data into a clean, structured format that machine learning algorithms can effectively learn from. Option B correctly identifies this as handling nulls, scaling numerical features, and encoding categorical variables, which are essential because ML models require numeric input and are sensitive to missing values and feature magnitudes.

Answer

Encrypting sensitive data before storing it in Azure for security compliance

Answer

The process of splitting raw data into training and test sets

Answer

Compressing data files to reduce the cost of Azure Blob Storage

Question 4

Which responsible AI principle requires that AI systems have mechanisms for people to raise concerns and seek redress?

Accepted Answer

Accountability. The Accountability principle in responsible AI ensures that AI systems are designed with mechanisms for human oversight, feedback, and redress. This includes providing clear channels for users to raise concerns about system behavior and seek remedies for any harm caused. Microsoft's responsible AI framework explicitly ties accountability to the ability to audit, review, and contest AI decisions.

Answer

Transparency

Answer

Reliability

Answer

Fairness

Question 5

What type of AI workload involves training a model to play games by rewarding successful moves?

Accepted Answer

Reinforcement learning where the agent receives rewards for successful moves. Reinforcement learning is the correct AI workload because it involves an agent learning to make decisions by interacting with an environment and receiving rewards or penalties for its actions. In game-playing scenarios, the model is trained through trial and error, where successful moves are rewarded, guiding the agent to maximize cumulative reward over time.

Answer

Supervised learning with labeled game states

Answer

Clustering similar game strategies together

Answer

Regression to predict the final game score

Question 6

What is 'knowledge mining' as an AI workload?

Accepted Answer

Using AI to extract meaningful information and insights from large volumes of unstructured content. Knowledge mining is an AI workload that uses services like Azure Cognitive Search to extract structured insights from unstructured data (documents, images, audio). It applies built-in AI skills (e.g., OCR, entity recognition, key phrase extraction) to index and enrich content, enabling search and discovery of hidden patterns.

Answer

Mining cryptocurrency using AI-optimized algorithms

Answer

A technique for extracting rare earth minerals used in GPU manufacturing

Answer

Automatically generating training data from existing knowledge bases

Question 7

What is 'model interpretability' and why is it important in responsible AI?

Accepted Answer

Understanding and explaining why a model produces specific predictions to enable trust and auditing. Model interpretability refers to the ability to understand and explain why a model produces specific predictions. It is a critical component of responsible AI because it enables trust, accountability, and auditing by allowing stakeholders to verify that decisions are fair, unbiased, and based on relevant features rather than spurious correlations.

Answer

The ability to translate a model's code into multiple programming languages

Answer

The speed at which a model processes inference requests

Answer

The accuracy of a model as measured on a standard benchmark dataset

Question 8

Which Azure AI capability can analyze video to identify and track specific people or objects across frames?

Accepted Answer

Azure AI Video Indexer. Azure AI Video Indexer is the correct choice because it is specifically designed to analyze video content, including the ability to detect, track, and identify people or objects across frames using AI-powered computer vision and audio analysis. It provides features like face detection, object tracking, and motion detection over time, making it suitable for this scenario.

Answer

Azure AI Custom Vision

Answer

Azure AI Face

Answer

Azure AI Vision OCR

Question 9

A robotic arm in a factory needs to pick parts from a bin. The system must identify each part and its exact outline to ensure precise grasping. Which Computer Vision capability should be used?

Accepted Answer

Semantic segmentation. Semantic segmentation is the correct capability because it classifies each pixel in an image, providing a precise outline of each part. This pixel-level classification is essential for a robotic arm to determine the exact shape and boundaries of parts for accurate grasping, unlike object detection which only provides bounding boxes.

Answer

Object detection

Answer

Image classification

Answer

Optical Character Recognition

Question 10

What is 'regression' in machine learning and when is it used?

Accepted Answer

Predicting a continuous numerical value such as price, temperature, or demand. Regression is a supervised machine learning technique used to predict a continuous numerical value, such as price, temperature, or demand, based on input features. It models the relationship between independent variables and a dependent variable that has a real-valued output, making option B correct.

Answer

A model that predicts which category an item belongs to from a set of options

Answer

Going back to a previous model version when the current version performs poorly

Answer

A technique for reducing the dimensionality of training data before model fitting

Question 11

What is Azure AI Language's text summarization capability used for?

Accepted Answer

Condensing long text into shorter summaries capturing the key information. Azure AI Language's text summarization capability is designed to condense long documents into shorter summaries that capture the key information. It uses extractive or abstractive summarization techniques to identify and present the most important sentences or generate new concise text, making it ideal for quickly digesting large volumes of content.

Answer

Translating long documents into multiple languages

Answer

Generating new creative text based on document themes

Answer

Classifying documents into predefined business categories

Question 12

Which responsible AI principle focuses on protecting personal information and ensuring AI systems handle data with appropriate privacy safeguards?

Accepted Answer

Privacy and security. Privacy and security is the correct responsible AI principle because it directly addresses the protection of personal data and the implementation of safeguards such as encryption, access controls, and data minimization. In AI systems, this principle ensures that sensitive information (e.g., PII) is handled in compliance with regulations like GDPR and that models do not inadvertently leak training data through inference attacks.

Answer

Fairness

Answer

Inclusiveness

Answer

Accountability

Question 13

A marketing team uses Azure OpenAI Service to generate product descriptions. They want the descriptions to follow a specific brand voice (formal, concise) and avoid generating any harmful or offensive language. Which combination of features should the team use?

Accepted Answer

A: Fine-tune the model with brand-specific data and enable content filtering.. Fine-tuning the model with brand-specific data allows the model to learn the desired brand voice (formal, concise) by adjusting its weights based on a curated dataset. Enabling content filtering ensures that any harmful or offensive language is blocked, either by Azure's built-in content moderation or by custom filters, meeting the safety requirement. This combination directly addresses both the style and safety needs.

Answer

B: Use few-shot learning with examples and disable content filtering for creativity.

Answer

C: Increase the temperature parameter and use the logprobs parameter.

Answer

D: Use the top_p parameter and set max_tokens to a low value.

Question 14

What are 'embeddings' in Azure OpenAI and what are they used for?

Accepted Answer

Numerical vector representations of text that capture semantic meaning for search and similarity tasks. Embeddings in Azure OpenAI are numerical vector representations of text that capture semantic meaning, enabling tasks like semantic search, clustering, and similarity comparisons. They convert words, sentences, or documents into high-dimensional vectors so that similar meanings are represented by vectors close to each other in the vector space. This is correct because embeddings are fundamental to modern AI search and recommendation systems, not related to hardware or web embedding tags.

Answer

Embedded systems software that runs AI models on IoT devices

Answer

HTML embed tags for displaying AI model outputs in web applications

Answer

Compressed versions of large language models that use fewer parameters

Question 15

What is tokenization in the context of natural language processing?

Accepted Answer

Breaking text into smaller units (tokens) for processing by language models. Tokenization is the process of breaking text into smaller units called tokens, which can be words, subwords, or characters. This is a fundamental preprocessing step in natural language processing because language models like GPT or BERT operate on discrete tokens rather than raw text. Each token is mapped to an integer ID from a vocabulary, enabling the model to process and generate language mathematically.

Answer

Encrypting text for secure transmission to AI services

Answer

Assigning security access tokens to API calls

Answer

Converting text into a numerical score for sentiment analysis