Microsoft Azure AI Fundamentals Practice Test (AI-900)

The correct answer is Optical Character Recognition (OCR). OCR is the technology used to extract printed or handwritten text from images, such as scanned documents like invoices. This allows the text to be converted into a machine-readable format, which is essential for automating data entry. Object detection is used to identify the location of objects in an image. Image classification is used to categorize an entire image. Facial detection is used to locate human faces in an image.

In machine learning, the label is the variable that represents the output or the value the model is trying to predict. In this scenario, the model aims to predict the 'Final Selling Price' of products, making it the label.

The other variables such as 'Production Cost', 'Marketing Budget' and 'Competitor Prices' are features that the model uses to make the prediction.

Generative models can produce new content such as marketing slogans by learning from existing data. They can generate creative and brand-aligned slogans that resonate with the target audience, making this approach appropriate.

Deep Learning techniques excel at processing large amounts of unstructured data like images and text because they can automatically learn hierarchical representations and features directly from the data without manual feature engineering. Other techniques like logistic regression, k-means clustering, and decision trees are generally more effective on structured data and may require significant feature engineering.

The Codex models in Azure OpenAI Service are specifically designed to translate natural language prompts into code across various programming languages, making them ideal for automating code creation tasks.

The GPT-3 models are geared towards generating human-like text but are not specialized for code.

The DALL·E models generate images from textual descriptions, not code.

The Embeddings models are used for finding text similarities and are not suitable for generating code from descriptions.

Computer Vision focuses on processing and interpreting visual information from images or videos. It is the most suitable choice for analyzing video surveillance to detect suspicious activities.

Natural Language Processing deals with understanding and generating human language, which is not applicable to visual data.

Knowledge Mining involves extracting insights from large volumes of structured and unstructured data, typically text-based.

Generative AI is concerned with creating new content rather than analyzing existing footage.

Object Detection is the appropriate technique because it not only identifies instances of objects within an image but also provides their locations, typically using bounding boxes. This makes it suitable for locating and counting products on shelves.

Image Classification assigns a label to an entire image but does not identify individual objects within the image or their positions.

Optical Character Recognition (OCR) is used to extract text from images, which is not relevant for detecting products.

Semantic Segmentation classifies each pixel in an image, assigning a class label to every pixel, which can provide more detailed localization but is more complex than necessary for simply locating and counting objects.

A model specialized in code generation - This is the correct answer. For generating code snippets from natural language descriptions, the developer should use a model like Codex, which is specifically designed for code generation. It can understand natural language inputs and generate corresponding code in various programming languages.

A model specialized in image generation - This type of model, like DALL·E, is designed for generating images from text descriptions, not for generating code.

A model specialized in sentiment analysis - Sentiment analysis models are used to assess the emotional tone of text, not to generate code based on natural language descriptions.

A model specialized in generating long-form text - While models like GPT are capable of generating text, they are not specifically optimized for code generation, making them less ideal for this task compared to a model like Codex.

Azure OpenAI Service offers advanced language models that can generate human-like text based on input prompts, making it suitable for content creation and code suggestions. The other services listed do not provide text generation capabilities.

Azure Translator Service is designed for translating text between languages.

Azure Cognitive Search enables indexing and querying of content.

Azure Text Analytics provides features like sentiment analysis and key phrase extraction but does not generate new text.

In machine learning, the label is the target variable that the model is intended to predict. Here, 'The price of the property' is what the model aims to predict, making it the label.

The other options are input variables, or features, that provide information to help the model make accurate predictions.

Image Classification assigns a single label to an entire image by analyzing its overall content. This differs from object detection, which identifies and locates individual objects within an image, and image segmentation, which classifies each pixel to understand the image at a granular level.

Object Detection is the appropriate solution because it not only recognizes objects within an image but also determines their locations by drawing bounding boxes around them. This allows the system to both identify damaged packages and pinpoint where they are in each image.

Image Classification could determine if an image contains a damaged package but cannot locate it within the image.

Optical Character Recognition (OCR) is used for extracting and interpreting text from images, which is not applicable in this scenario.

Facial Detection focuses on identifying human faces and would not be useful for detecting packages.

Azure Machine Learning Endpoints allow you to deploy trained machine learning models as web services, enabling real-time predictions through RESTful APIs.

Azure Machine Learning Pipelines are used to build and manage automated workflows.

Azure Machine Learning Compute Instances provide development environments for machine learning tasks.

Azure Blob Storage is a service for storing unstructured data.

They create new data based on learned patterns from training data - This is the correct answer. AI models that generate new content, such as text, images, or music, are typically generative models. They learn patterns from training data and use this knowledge to produce original content that resembles the data they were trained on.

They classify input data into predefined categories - This describes classification models, which focus on categorizing data into specific labels or categories, rather than generating new content.

They predict numerical values from historical data - This is characteristic of predictive models, which are used for forecasting or regression tasks, rather than generating new content.

They retrieve exact copies of existing content - This describes retrieval-based models, which focus on finding and returning pre-existing content from a database or dataset, rather than generating new, original content.

Generating personalized product descriptions uses a text-generation model that creates new, tailored content for each visitor based on patterns learned from their browsing data. This is a clear example of generative AI, which focuses on producing novel outputs rather than selecting from fixed templates.

The other options are not generative:

• A scripted chatbot simply retrieves predefined replies, so no new content is generated.
• Analyzing purchasing patterns to recommend items is predictive analytics, concerned with forecasting rather than content creation.
• Image recognition that categorizes uploads is a classification task in computer vision, not content generation.

Removing personally identifiable information (PII) from the data, also known as data anonymization, is the most effective way to protect customer privacy during model training.

While encrypting data and limiting access are important security measures, they do not prevent the AI model from potentially learning sensitive information. Using secure servers protects data from external threats but does not address privacy concerns related to data handling during training.

Leverage the service's ability to generate images from text descriptions - This is the correct answer. Azure OpenAI Service provides models like DALL·E, which can generate images based on descriptive text inputs.

Use code generation features to create image-rendering scripts - This feature is focused on generating code for specific tasks

Apply language translation capabilities to interpret user inputs - Language translation is used for converting text from one language to another

Implement image analysis to extract information from images - Image analysis focuses on understanding and processing images

Facial detection and analysis solutions locate each face in an image and can return attributes such as bounding-box coordinates, head pose, and a limited age estimate. These details let the marketing team derive per-person demographic insights.

Image classification assigns one or more labels to an entire image but does not return per-face information. Object detection draws boxes around many object types but does not calculate facial attributes like age. Optical character recognition (OCR) only extracts text, so it is irrelevant here.

The Azure AI Language service provides pre-built features for sentiment analysis, key phrase extraction, and named entity recognition, enabling developers to analyze text data with minimal custom development.

Azure Machine Learning is a platform for building and deploying custom machine learning models, which would require more effort to implement these features.

Azure AI Speech service focuses on speech processing, such as speech-to-text and text-to-speech, not text analytics.

Azure Cognitive Search is designed for indexing and searching over data but does not provide text analytics capabilities like sentiment analysis or entity recognition.

Image Classification - This is the correct answer. Image classification assigns labels to images based on the overall visual content, without pinpointing the location of specific objects. It categorizes the entire image into predefined classes for example "dog," "cat," "car" but it does not locate individual objects within the image.

Object Detection involves identifying and locating specific objects within an image, often with bounding boxes, in addition to classifying the objects. This is different from image classification, which does not detect object locations.

Optical Character Recognition (OCR) is used to extract and recognize text from images, but it is not focused on classifying images or identifying overall content.

Semantic Segmentation divides an image into regions that correspond to different object categories and labels every pixel in the image, which is more detailed than image classification. It assigns labels to every part of the image, rather than classifying the entire image at once.