KI-Glossar
Das vollständige Wörterbuch der Künstlichen Intelligenz
Data Version Control (DVC)
Version control system for data and ML models that integrates with Git to track changes in datasets. DVC allows versioning large files, reproducing experiments, and collaborating efficiently on data science projects.
Amazon SageMaker
AWS managed platform that provides a complete environment for building, training, and deploying machine learning models at scale. SageMaker integrates tools for data labeling, distributed training, hyperparameter tuning, and serverless deployment.
Azure Machine Learning
Microsoft's cloud service offering an integrated environment for the complete machine learning lifecycle, from data preparation to model monitoring. Azure ML provides MLOps tools for automation, versioning, and CI/CD deployment of models.
Google Vertex AI
Google Cloud's unified machine learning platform that combines AutoML and AI Platform to simplify the development and deployment of ML models. Vertex AI offers tools for custom training, model monitoring, and prediction explanation.
ML Pipeline Orchestration
Process of automating and coordinating different steps of a machine learning workflow, from data ingestion to model deployment. Orchestration ensures sequential or parallel execution of tasks with dependency and error management.
Model Deployment Patterns
Standardized architectures and strategies for putting ML models into production, including real-time, batch, edge, or streaming deployment. These patterns define how models serve predictions while optimizing performance and costs.
MLOps Platform
Integrated software solution that provides a consistent set of tools to automate and orchestrate the entire machine learning lifecycle. These platforms unify model development, deployment, and monitoring in a controlled environment.
AutoML Platform
System that automates the process of building ML models, including feature selection, algorithm choice, and hyperparameter optimization. AutoML platforms allow non-experts to quickly develop high-performing models.
ML Model Governance
Set of policies, processes, and tools to ensure regulatory compliance, ethics, and transparency of machine learning models. Governance covers the complete lifecycle of models, from their development to their retirement.