Technical term in machine learning and statistics; refers to a model that fits training data too closely and fails to generalize. The Japanese synonym 過学習 is also common.
See also: 過学習
We apply regularization to prevent overfitting.
This model may be overfitting.