Overfitting: a problem that occurs when a model is trained too well on the training data and performs poorly on new, unseen data.