AI-900 Flashcards — Free Microsoft Azure AI Fundamentals AI-900 Study Cards

Fairness

Fairness is the correct principle because it directly addresses the need to prevent discrimination in AI systems, such as loan approval models. By focusing on fairness, the bank ensures that the model's predictions do not systematically disadvantage any group based on protected attributes like race, gender, or age, which is critical for ethical and legal compliance.

Regression

Predicting the exact market price of a house is a regression task because the target variable (price) is a continuous numeric value. Regression algorithms, such as linear regression or decision tree regression, learn the relationship between input features (e.g., square footage, bedrooms, location) and a continuous output. In Azure Machine Learning, you would select a regression model from the designer or AutoML to solve this problem.

Image classification

Image classification assigns a single label to an entire image based on its dominant content. Since the requirement is to output one label (car, truck, or motorcycle) per image, this maps directly to Azure's Custom Vision image classification capability, which trains a model to categorize whole images into predefined classes.

Text Analytics

Text Analytics (now part of Azure AI Language) is the correct service because it provides pre-built entity extraction capabilities specifically designed to identify and extract named entities like people (patient names), quantities (medication dosages), and dates from unstructured text. This aligns directly with the requirement to extract specific data elements from doctors' notes without needing custom model training.

A. Azure OpenAI Service

Azure OpenAI Service provides access to large language models (LLMs) like GPT-4, which are specifically designed for generative tasks such as creating unique, human-like text from a few input keywords. This makes it the ideal choice for automatically generating product descriptions at scale, as it can produce varied and contextually relevant content without requiring pre-labeled training data.

Transparency

The system's inability to explain how it reached a decision violates the transparency principle, which requires AI systems to be understandable and interpretable. Complex neural networks often act as black boxes, making it impossible to provide meaningful explanations to users, directly contradicting Microsoft's responsible AI guideline that decisions should be explainable.

Tracking each iteration of a trained model for rollback, A/B testing, auditing, and reproducibility

Model versioning is the practice of tracking each iteration of a trained model, including its hyperparameters, training data snapshot, and evaluation metrics. In MLOps, it is essential because it enables rollback to a previous model if a new version performs poorly, supports A/B testing by comparing multiple model versions in production, provides an audit trail for compliance, and ensures reproducibility by capturing the exact code, data, and environment used to train each version.

Ensuring people understand when they're interacting with AI, how it works, and what its limitations are

Option B is correct because AI transparency, as defined in Microsoft's Responsible AI principles, is about ensuring that users understand when they are interacting with an AI system, how the system makes decisions, and what its limitations are. This principle focuses on clear communication and documentation, not on open-sourcing code or financial reporting.

Increase the frequency_penalty parameter.

Increasing the frequency_penalty parameter reduces the likelihood of the model repeating the same phrases by penalizing tokens that have already appeared in the generated text. This directly addresses the repetitive loops without altering the overall creativity, as frequency_penalty specifically targets token frequency rather than randomness or diversity.

Root Mean Squared Error (RMSE)

Root Mean Squared Error (RMSE) is the correct metric because it measures the average magnitude of prediction errors in the same units as the target variable (price). RMSE is computed as the square root of the average squared differences between predicted and actual values, which brings the error metric back to the original unit (e.g., dollars), making it directly interpretable for regression tasks like house price prediction.

Content filtering

Content filtering in Azure OpenAI Service uses a set of pre-built, configurable filters to detect and block harmful content categories such as hate, violence, sexual, and self-harm. This feature operates at the service level, intercepting both prompts and completions to prevent offensive language or harmful stereotypes from being generated, regardless of how the prompt is phrased.

Semantic Segmentation

Semantic segmentation assigns a class label to every pixel in an image, making it the correct choice for precisely delineating tumor boundaries in MRI scans. Azure Computer Vision's semantic segmentation capability outputs a pixel-level mask, enabling the research team to differentiate tumor, healthy tissue, and background at the finest granularity.

Frequency penalty

Frequency penalty directly reduces repetition by penalizing tokens that have already appeared in the generated text. A higher frequency penalty value (e.g., 0.5 to 1.0) decreases the likelihood of the model reusing the same phrases or words, making the output more diverse and less repetitive.

Data imbalance

The model achieves 98% accuracy by simply predicting all emails as 'not spam', which reflects the 95% majority class in the dataset. This is a classic symptom of class imbalance, where the model learns to exploit the skewed distribution rather than learning meaningful patterns to distinguish spam from non-spam. In Azure Machine Learning, techniques like SMOTE or stratified sampling are used to mitigate this issue.

Set temperature to 0

Setting temperature to 0 forces the model to choose the most likely token at each step, eliminating randomness and producing deterministic outputs for the same prompt. This is essential for reproducibility in testing scenarios where identical results are required across API calls.

Recall

Recall (also known as sensitivity) measures the proportion of actual positive cases correctly identified by the model. In this scenario, the model predicts all transactions as legitimate, so it correctly identifies zero fraudulent transactions, giving a recall of 0%. Accuracy alone is misleading because the dataset is highly imbalanced (99.5% legitimate, 0.5% fraudulent), and a 99.5% accuracy can be achieved by simply predicting the majority class. Recall directly reveals the model's failure to detect any fraud.

Frequency_penalty

The frequency_penalty parameter reduces the likelihood of the model repeating the same phrases by penalizing tokens that have already appeared in the generated text. A higher frequency_penalty value (e.g., 0.5 to 1.0) discourages the model from reusing the same words or phrases, making the output more diverse and natural. This directly addresses the issue of repetitive phrasing in product reviews.

On average, the model's predictions are $5,000 away from the actual prices.

Option A is correct because Mean Absolute Error (MAE) measures the average absolute difference between predicted and actual values. An MAE of $5,000 means that, on average, each prediction deviates from the true house price by $5,000. This is a standard interpretation of MAE in regression metrics.

Overfitting

The model achieved an R-squared of 0.95 on the test set, indicating excellent performance on data from the same distribution. However, when deployed to a new city with different property characteristics, the predictions were very inaccurate. This is a classic symptom of overfitting, where the model has learned noise and patterns specific to the training data (e.g., city-specific price trends) that do not generalize to unseen data from a different distribution.

Optical Character Recognition (OCR)

Option A is correct because Azure Computer Vision's Optical Character Recognition (OCR) API is specifically designed to extract printed or handwritten text from images. In this scenario, the handwritten shipping labels contain textual data (destination address, sender name, package weight), and OCR can read and digitize that text for automated processing. The other options address different visual tasks—object detection, classification, or facial recognition—none of which extract text content.

Custom Vision object detection

Custom Vision object detection is the correct choice because it not only classifies images (defective vs. non-defective) but also localizes defects by drawing bounding boxes around them. The labeled dataset with defect locations directly supports training a model to output both class labels and spatial coordinates, which is exactly what object detection provides.

Conversational Language Understanding (CLU)

Conversational Language Understanding (CLU) is the correct Azure AI Language feature because it is specifically designed to extract both intents (e.g., 'Book a flight') and entities (e.g., destination, date) from user utterances. The developer has a small set of labeled examples, which CLU can use to train a custom model for intent recognition and entity extraction, making it ideal for building a virtual assistant.

Optical Character Recognition (OCR)

Optical Character Recognition (OCR) is the correct choice because it is specifically designed to extract printed or handwritten text from images, including numbers and special characters. Azure Computer Vision's OCR API can handle varied handwriting styles and convert them into machine-readable text, making it ideal for transcribing historical artifact labels.