Stationarity is the property of a feature, process, or distribution remaining stable across one or more dimensions, commonly time. In machine learning, stationarity supports stronger assumptions about generalization, while nonstationarity can signal drift or changing conditions.
A feature whose values don't change across one or more dimensions, usually time. For example, a feature whose values look about the same in 2021 and 2023 exhibits stationarity. In the real world, very few features exhibit stationarity. Even features synonymous with stability (like sea level) change over time. Contrast with nonstationarity.