0 alternative definitions
The degradation of a machine learning model’s performance over time due to changes in data distributions that differ from those used in training.