Technical term in machine learning and statistics. Refers to a model that fits the training data too closely and fails to generalize. Often contrasted with 過学習 (overlearning).
See also: 過学習
We apply regularization to prevent overfitting.
This model is likely suffering from overfitting.