# Outliers

> Values distant from most other values.

Values distant from most other values.

## Machine learning

In [machine learning](https://wiki.g15e.com/pages/Machine%20learning.txt), any of the following are outliers:[^1]

- Input data whose values are more than roughly 3 [standard deviations](https://wiki.g15e.com/pages/Standard%20deviation.txt) from the <mean>.
- [weights](https://wiki.g15e.com/pages/Weight%20(machine%20learning.txt)) with high absolute values.
- Predicted values relatively far away from the actual values.

Outliers are often caused by typos or other input mistakes. In other cases, outliers aren't mistakes; after all, values five standard deviations away from the mean are rare but hardly impossible.[^1]

Outliers often cause problems in model training. [Clipping](https://wiki.g15e.com/pages/Clipping%20(machine%20learning.txt)) is one way of managing outliers.[^1]

## Footnotes

[^1]: https://developers.google.com/machine-learning/glossary#outliers