AI-900 Practice Test 50 — 15 Questions

Question 1

What is 'stochastic gradient descent' (SGD) and how does it work?

Accepted Answer

An optimisation algorithm that updates weights using gradients computed on random data mini-batches. Stochastic Gradient Descent (SGD) is an optimization algorithm used to train machine learning models by iteratively updating model weights. It computes the gradient of the loss function on a randomly selected mini-batch of training data (not the entire dataset), which introduces noise but significantly speeds up convergence and reduces memory usage. This mini-batch approach is the core of SGD and distinguishes it from batch gradient descent.

Answer

A random sampling method for selecting training data without replacement

Answer

A technique for randomly selecting which model architecture to use for AutoML

Answer

Randomly descending through decision tree branches to make predictions

Question 2

A marketing team wants to create original images for advertisements based on text descriptions. Which Azure OpenAI Service model capability should they use?

Accepted Answer

DALL-E. DALL-E is the correct choice because it is the Azure OpenAI Service model specifically designed for generating original images from natural language text descriptions. Unlike other models in the suite, DALL-E uses a diffusion-based architecture to create photorealistic or stylized visuals based on prompt inputs, making it ideal for the marketing team's goal of producing custom advertisement imagery.

Answer

GPT-3.5

Answer

Codex

Answer

Azure Speech-to-Text

Question 3

What is 'online learning' (incremental learning) in machine learning?

Accepted Answer

Continuously updating model weights on new data as it arrives rather than batch retraining. Online learning (incremental learning) is a machine learning technique where the model is updated continuously as new data arrives, rather than retraining from scratch on the entire dataset. This is essential for scenarios with streaming data or when retraining on all historical data is computationally prohibitive. In Azure, this is supported by services like Azure Stream Analytics and Azure Machine Learning's online endpoints, which can update model weights incrementally.

Answer

Training ML models through an online learning management system

Answer

Requiring an internet connection during model training for cloud compute access

Answer

A training approach where users can interact with and correct the model in real time

Question 4

What is a copilot in the context of Microsoft AI products?

Accepted Answer

An AI assistant integrated into products that helps users complete tasks using natural language. A copilot in Microsoft AI products, such as Microsoft 365 Copilot or GitHub Copilot, is an AI assistant that uses large language models (LLMs) and natural language processing to help users complete tasks like drafting documents, generating code, or summarizing emails. It integrates directly into the user interface of applications (e.g., Word, Excel, Teams) and interprets natural language prompts to produce contextually relevant outputs, leveraging Azure OpenAI Service under the hood.

Answer

A hardware accelerator for AI model training

Answer

A type of database for storing conversation history

Answer

A software testing tool for AI models

Question 5

What is the difference between extractive summarization and abstractive summarization?

Accepted Answer

Extractive pulls existing sentences; abstractive generates new text capturing the meaning. Extractive summarization identifies and extracts the most important sentences directly from the source text, while abstractive summarization generates new sentences that capture the core meaning, often using natural language generation techniques. In Azure AI Language, extractive summarization returns a set of ranked sentences from the original document, whereas abstractive summarization produces a concise summary that may rephrase content. This distinction is fundamental to understanding how different NLP workloads handle text summarization tasks.

Answer

Extractive works on text; abstractive works on images

Answer

Extractive is for long documents; abstractive is for short text

Answer

Extractive summarization is always less accurate than abstractive

Question 6

What is 'explainable AI' (XAI) and why is it important for responsible AI?

Accepted Answer

Techniques that make AI decision-making understandable to humans, supporting transparency and trust. Explainable AI (XAI) refers to a set of techniques and methods that produce human-understandable explanations of AI model decisions, outputs, and behaviors. It is critical for responsible AI because it enables transparency, builds user trust, supports regulatory compliance (e.g., GDPR's right to explanation), and helps identify and mitigate bias or errors in model predictions.

Answer

AI systems that can explain jokes and riddles to users

Answer

AI systems designed to teach other AI systems

Answer

AI that automatically generates explanations of its errors

Question 7

What is the primary use case for DALL-E models available in Azure OpenAI?

Accepted Answer

Generating images from text descriptions. DALL-E models are specifically designed for generative image creation, taking natural language text descriptions as input and producing corresponding images. In Azure OpenAI, this capability is exposed through the DALL-E API, which uses a transformer-based architecture trained on image-text pairs to generate novel visual content from prompts. This makes option B the correct answer because it directly matches the primary use case of DALL-E: text-to-image generation.

Answer

Generating text responses to questions

Answer

Transcribing spoken audio to text

Answer

Detecting objects in photographs

Question 8

What is 'hallucination' in large language models and what techniques help reduce it?

Accepted Answer

When a model generates confident but factually incorrect or fabricated information. Hallucination in large language models (LLMs) refers to the generation of text that is confident, coherent, and plausible-sounding but factually incorrect or entirely fabricated. This occurs because LLMs are probabilistic next-token predictors trained on vast datasets, not databases of verified facts; they lack a built-in mechanism to distinguish truth from fiction. Techniques to reduce hallucination include grounding outputs with retrieval-augmented generation (RAG) using Azure AI Search, prompt engineering with system messages that constrain responses to verified sources, and fine-tuning with human feedback (RLHF) to penalize factual errors.

Answer

When a model generates images instead of text in response to a text prompt

Answer

When users imagine the AI is sentient due to very convincing responses

Answer

When a model's training data contains copyrighted material it memorises

Question 9

A developer is using Azure OpenAI with GPT-4 to build a chatbot that answers legal questions based on a company's internal policy documents. The developer wants the model's responses to be maximally deterministic and factual, avoiding any creative or speculative language. Which parameter should the developer set to the lowest possible value in the API call?

Accepted Answer

Temperature. Temperature controls the randomness of the model's output. Setting it to the lowest possible value (0) makes the model deterministic, always choosing the most likely next token, which is ideal for factual, non-creative responses like legal answers. Higher temperature values introduce variability and creativity, which would be undesirable for this use case.

Answer

Frequency penalty

Answer

Presence penalty

Answer

Top_p

Question 10

What is model monitoring in Azure Machine Learning and why is it important?

Accepted Answer

Tracking model prediction quality and data distribution changes in production to detect degradation. Model monitoring in Azure Machine Learning is the continuous tracking of a deployed model's performance in production, focusing on prediction quality (e.g., accuracy, precision, recall) and data distribution shifts (data drift) to detect degradation over time. This is critical because models can become stale as real-world data evolves, leading to poor business decisions or compliance failures. Azure ML's Model Data Collector and monitoring dashboards automatically capture input data and predictions, alerting data scientists when drift or performance drops below defined thresholds.

Answer

Checking how many API calls the model endpoint receives per hour

Answer

Monitoring the GPU memory usage during model training

Answer

Reviewing model architecture choices for optimization

Question 11

What is cross-validation in machine learning?

Accepted Answer

Repeatedly training and evaluating the model on different data splits for reliable performance estimates. Cross-validation is a technique for assessing how a machine learning model will generalize to an independent dataset. It involves partitioning the data into complementary subsets, training the model on one subset (the training fold), and validating it on the remaining subset (the validation fold), then repeating this process multiple times with different partitions. The final performance estimate is the average of the validation scores, which provides a more reliable and less biased measure than a single train-test split.

Answer

Training multiple different models and comparing their performance

Answer

Checking if a model works correctly by running it backward

Answer

Training a model on two different datasets simultaneously

Question 12

A company is developing an AI voice assistant for children. The assistant must respond with age-appropriate language and avoid providing any harmful instructions. Which Microsoft responsible AI principle is most directly relevant to ensuring the system operates safely?

Accepted Answer

Reliability and Safety. The scenario requires the AI voice assistant to avoid providing harmful instructions and to operate safely for children. Microsoft's Reliability and Safety principle directly addresses the need for systems to function reliably and to fail safely, preventing harm even when unexpected inputs occur. This principle ensures that the assistant's responses are vetted for safety, which is the most relevant consideration here.

Answer

Fairness

Answer

Privacy and Security

Answer

Inclusiveness

Question 13

What is the purpose of a test dataset in machine learning model development?

Accepted Answer

To provide an unbiased final evaluation of the trained model on unseen data. The test dataset is used to provide an unbiased final evaluation of the trained model on unseen data. This is critical in machine learning because the model has never seen the test examples during training or validation, so the evaluation metrics (e.g., accuracy, precision, recall) reflect the model's true generalization ability. In Azure Machine Learning, the test dataset is typically split from the original data before any training begins and is only used once at the end of the model development lifecycle.

Answer

To provide additional examples for training the model

Answer

To tune hyperparameters and select the best model version

Answer

To monitor model performance after deployment

Question 14

What is 'document processing' as an AI workload and what pipeline does it typically involve?

Accepted Answer

Automating extraction, understanding, and routing of business documents through OCR, extraction, and NLP. Document processing as an AI workload involves automating the extraction, understanding, and routing of information from documents. This pipeline typically uses Optical Character Recognition (OCR) to digitize text, followed by AI models (e.g., Azure Form Recognizer) for data extraction, and Natural Language Processing (NLP) for semantic understanding and classification. Option B correctly captures this end-to-end automation, which is a core AI workload in Azure.

Answer

Using Azure Blob Storage to store and manage document files efficiently

Answer

Digitising physical documents by scanning them and converting to PDF format

Answer

Managing document access permissions and version control in SharePoint

Question 15

What is 'AI accountability' in Microsoft's Responsible AI principles?

Accepted Answer

Humans remaining responsible for AI systems with oversight mechanisms and clear lines of accountability. Microsoft's Responsible AI principle of accountability means that humans are ultimately responsible for AI systems. This includes establishing oversight mechanisms, clear lines of accountability, and ensuring that AI systems are designed and operated under human control. It does not refer to billing, automatic self-correction, or vendor liability.

Answer

Billing accountability — ensuring costs are tracked and charged to the correct Azure subscription

Answer

AI systems reporting their own mistakes and triggering automatic self-correction

Answer

Holding AI vendors legally accountable for damages caused by their models