Overfitting
Warning
This post is more than a year old. Information may be outdated.
Overfitting is a common problem in machine learning where a model learns the training data too well, including its noise and outliers, but does not generalize well to new, unseen data.