AI Dictionary
Overfitting
Definition
A modeling error that occurs when a function is too closely fit to a limited set of data points.
Deep Dive
Overfitting is a common and critical modeling error in machine learning where a model learns the training data too well, capturing not only the underlying patterns but also the noise and random fluctuations present in that specific dataset. This results in a model that performs exceptionally well on the data it was trained on but fails to generalize effectively to new, unseen data. Essentially, the model becomes too specific to its training examples, losing its ability to make accurate predictions or classifications on data it hasn't encountered before.
Examples & Use Cases
- 1A sales prediction model that perfectly forecasts past sales figures but is wildly inaccurate for future quarters
- 2A spam filter that's so specific it flags legitimate emails with minor variations as spam
- 3An image classifier that only recognizes the exact resolution and lighting of images it was trained on
Related Terms
UnderfittingGeneralizationRegularization