Definition

A modeling error that occurs when a machine learning model learns the training data too well, including its noise and outliers, resulting in poor performance on unseen data. Overfitting indicates the model has memorized rather than generalized.

Defined Term