Eine Pipeline ist eine geordnete Abfolge von Verarbeitungsschritten, die Daten oder Artefakte durch einen KI- oder Machine-Learning-Workflow bewegt. Sie kann Phasen wie Ingestion, Transformation, Training, Evaluation, Deployment und Monitoring umfassen.
All of the operations needed to fit a model to a data set. A pipeline consists of data import, transformation, featurization, and learning steps. Once a pipeline is trained, it turns into a model.
The infrastructure surrounding a machine learning algorithm. A pipeline includes gathering the data, putting the data into training data files, training one or more models, and exporting the models to production. See ML pipelines in the Managing ML Projects course for more information.