Microsoft Azure AI Fundamentals General 4 — Questions and Answers
Question 1: Which Azure service allows developers to add pre-built AI capabilities to applications without needing machine learning expertise?
- Azure Machine Learning
- Azure Databricks
- Azure AI Services (Cognitive Services) (Correct answer)
- Azure Synapse Analytics
Correct answer: Azure AI Services (Cognitive Services)
Azure AI Services (formerly Cognitive Services) offers pre-built APIs for vision, language, speech, and decision tasks that developers can integrate without ML expertise.
Question 2: What is 'overfitting' in a machine learning model?
- The model performs well on training data but poorly on new, unseen data (Correct answer)
- The model is too simple to capture patterns in the training data
- The model requires too much compute to train
- The model takes too long to make predictions
Correct answer: The model performs well on training data but poorly on new, unseen data
Overfitting occurs when a model memorizes training data so well that it fails to generalize to new data, resulting in low training error but high validation error.
Question 3: In Azure AI Vision, which feature identifies the age, gender, and emotion of people in photographs?
- Image tagging
- Optical Character Recognition (OCR)
- Face detection and analysis (Correct answer)
- Scene understanding
Correct answer: Face detection and analysis
Azure AI Vision's face detection and analysis feature can locate faces in images and return attributes such as estimated age, inferred gender, and detected emotion.
Question 4: Which Azure feature enables you to detect anomalies in time-series data such as server metrics or sensor readings?
- Azure AI Translator
- Azure Anomaly Detector (Correct answer)
- Azure AI Content Safety
- Azure AI Document Intelligence
Correct answer: Azure Anomaly Detector
Azure Anomaly Detector is an AI service that automatically identifies unexpected patterns in time-series data without requiring labeled training data.
Question 5: What does the term 'neural network' refer to in artificial intelligence?
- A network of physical servers in a data center
- A computational model inspired by the structure of the human brain (Correct answer)
- A set of rules programmed by developers
- A database of labeled training examples
Correct answer: A computational model inspired by the structure of the human brain
A neural network is a computational model made up of interconnected nodes (neurons) organized in layers, loosely inspired by the biological structure of the human brain.
Question 6: Which Azure AI service can extract structured data from forms, invoices, and receipts?
- Azure AI Language
- Azure AI Speech
- Azure AI Document Intelligence (Correct answer)
- Azure AI Personalizer
Correct answer: Azure AI Document Intelligence
Azure AI Document Intelligence (formerly Form Recognizer) uses AI to extract structured fields, tables, and key-value pairs from documents like invoices and receipts.
Question 7: What is the primary goal of 'feature engineering' in a machine learning workflow?
- Reducing the size of the training dataset
- Transforming raw data into informative input variables that improve model performance (Correct answer)
- Selecting the best machine learning algorithm
- Deploying the model to a production endpoint
Correct answer: Transforming raw data into informative input variables that improve model performance
Feature engineering involves creating, selecting, and transforming variables from raw data to provide the model with more informative inputs that improve predictive accuracy.
Microsoft Azure AI Fundamentals General 4
Which Azure service allows developers to add pre-built AI capabilities to applications without needing machine learning expertise?