AI-900 Certification Cheat Sheet

Azure Machine Learning's automated feature allowing users to build and deploy machine learning models without extensive coding.

Azure provides various tools and services specifically designed for data scientists to implement machine learning solutions.

MLOps is a practice for the management of machine learning models throughout their lifecycle, from development to deployment and monitoring.

Azure AI Vision is a suite of services within Azure that empowers developers to integrate powerful computer vision capabilities into their applications.

Generative AI refers to a type of artificial intelligence capable of creating new content, including images, text, audio, and code. It learns from existing data and generates novel outputs.

Azure OpenAI Service is a platform offered by Microsoft that provides access to powerful and advanced AI models, including GPT and DALL-E., enabling users to integrate these models into their applications.

Responsible AI practices ensure the development and deployment of AI systems that consider ethical implications, fairness, transparency, and accountability.

Azure OpenAI Service emphasizes responsible AI by incorporating features and guidelines to mitigate bias, promote transparency, and ensure ethical use.

The field of AI dedicated to enabling computers to understand and process human language.

Virtual assistants like Siri and Alexa use NLP to understand your voice commands and respond accordingly.

NLP powers translation services, bridging the gap between different languages and expanding global communication.

Through NLP, computers can analyze text to identify the underlying sentiment or emotion being expressed.

NLP enables chatbots to engage in natural conversations, providing assistance and answering questions in various contexts.

NLP algorithms can condense long documents into concise summaries, making it easier to grasp information quickly.

NLP techniques identify and categorize entities within text, aiding in information structuring and comprehension.

Human language is complex and ambiguous, posing challenges for NLP systems to accurately understand meaning.

Predicting a continuous target variable using a linear relationship between it and one or more input features.

A classification algorithm that predicts the probability of a data point belonging to a particular class based on a linear combination of features.

A powerful, versatile algorithm that uses a hyperplane to classify data points into different categories.

A tree-like structure that uses a series of decision nodes to divide data into smaller groups and predict the outcome.

An ensemble learning method that combines multiple decision trees to improve accuracy and generalization.

A probabilistic classification algorithm that uses Bayes' theorem and assumptions of feature independence.

A non-parametric method that classifies new data points based on their proximity to labeled data points.

A learning paradigm where algorithms discover patterns and relationships in unlabeled data without explicit guidance.

Datasets in Azure Machine Learning that contain labels alongside data samples.

A type of TabularDataset in Azure Machine Learning, specifically designed for data with labels. These datasets are exclusively generated from Azure Machine Learning data labeling projects.

Frameworks like TensorFlow or PyTorch used by developers to construct complex machine learning models. Accelerators like ONNX Runtime enhance their performance for faster training and inference.

The process of applying labels or annotations to data samples, providing ground truth information for training machine learning models.

A format for saving labeled data in Azure Machine Learning, commonly used for image data. It stores data references and their corresponding labels.

The process of extracting labels from a data labeling project and storing them in a format suitable for use in Azure Machine Learning.

A method used to load labeled datasets into a pandas DataFrame, enabling easy analysis and manipulation within Python.

A powerful tool for data analysis and manipulation in Python, offering methods for loading, exploring, and manipulating data in labeled datasets.

Microsoft's commitment to developing AI ethically, focusing on transparency, fairness, and accountability, with experts across research, policy, and engineering collaborating to ensure responsible AI deployment.

A framework that promotes effective and ethical AI development by outlining a set of practices and principles to guide AI development and deployment.

The process of systematically identifying potential negative impacts of AI systems before, during, and after deployment.

Establishing methods to track and assess the occurrence of potential harms identified in AI systems, using both manual and automated approaches.

Implementing strategies to reduce or eliminate identified harms from AI systems, focusing on mitigation measures at various levels.

A set of practices and principles that prioritize transparency and explainability in AI models, enabling users to understand and trust the decision-making process.

Automated machine learning (AutoML) streamlines model selection and hyperparameter tuning. Azure ML evaluates various algorithms and configurations to identify the most effective model.

A set of practices that streamline the development and deployment of machine learning models, improving efficiency and collaboration.

Azure ML allows you to deploy machine learning models as web services or within containers. You can monitor model performance, make adjustments, and manage model versions.

Azure AI's offering that provides access to powerful language models and APIs, enabling developers to build AI-powered applications.

Azure ML seamlessly integrates with other Azure services like Databricks, Data Factory, and Synapse Analytics. This allows for comprehensive machine learning workflows, leveraging data pipelines and coordination tools.

Predictive analytics uses Azure ML to build models for tasks like predicting sales, customer churn, and demand forecasting.

Azure ML incorporates pre-built AI capabilities like image recognition, optical character recognition (OCR), and sentiment analysis. It empowers developers to easily integrate these features into their applications.

Anomaly detection with Azure ML helps identify unusual patterns in various areas like manufacturing processes, network traffic, or financial transactions.

Recommendation systems, powered by Azure ML, create personalized recommendations for users by analyzing their behavior and preferences.

Azure ML employs strong security protocols to protect both data and trained models. It ensures compliance with privacy regulations and industry standards.