[ fee-cher set ] n. A feature set is a subset of available features that are selected to power an analytical function or model.
How Feature Sets Work in Molecula
Feature sets in Molecula are typically administered by the data engineer and used by the data scientist and/or application developers. These can be configured to serve online, offline, or analytical workloads for model serving, model training, and real-time, adHoc analytics. Each feature set has its own https endpoint, documentation, distinct security model, and allows for workload isolation in both development and production environments. Features in a feature set, can be a subset selected directly from a single feature table, can be derived from joins performed across multiple feature tables, and can be based on calculations and transformations performed on underlying features.NEXT