Glossary term
Glossary term
Foundations
The group of features your machine learning model trains on. For example, a simple feature set for a model that predicts housing prices might consist of postal code, property size, and property condition.
Created for this library
A retail forecasting team locks a stable feature set per model generation so production retraining stays reproducible.
A risk modeling team versions its feature set and documents each feature in a feature catalog for compliance review.
A churn team aligns its feature set with the operational data warehouse so production scoring can rely on the same fields used during training.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License