Loading…
Glossary term
When a model learns its training data too closely, including noise, and performs poorly on new data.
An overfit model memorizes its training examples, including their noise, instead of learning the general pattern. It scores very well on training data and poorly on anything new.
Warning sign: training error keeps falling while validation error starts to rise. Remedies include more data, a simpler model, regularization and stopping training early.
Overfitting is the most common reason a model that looked great in testing fails in the real world.