Mean absolute error

For a broader coverage related to this topic, see Mean • Mean percentage error
absolute difference.
• Symmetric mean absolute percentage error

In statistics, the mean absolute error (MAE) is a quan-

tity used to measure how close forecasts or predictions
are to the eventual outcomes. The mean absolute error is
3 References
given by
[1] “2.5 Evaluating forecast accuracy | OTexts”. www.otexts.
org. Retrieved 2016-05-18.

1∑ 1∑
n n
[2] Hyndman, R. and Koehler A. (2005). “Another look at
MAE = |fi − yi | = |ei | . measures of forecast accuracy”
n i=1 n i=1

As the name suggests, the mean absolute error is an av-

erage of the absolute errors |ei | = |fi − yi | , where fi is
the prediction and yi the true value. Note that alternative
formulations may include relative frequencies as weight
factors.
The mean absolute error used the same scale as the data
being measured. This is known as a scale-dependent ac-
curacy measure and therefore cannot be used to make
comparisons between series using different scales.[1]
The mean absolute error is a common measure of forecast
error in time series analysis,[2] where the terms “mean
absolute deviation” is sometimes used in confusion with
the more standard definition of mean absolute deviation.
The same confusion exists more generally.

1 Related measures
The mean absolute error is one of a number of ways of
comparing forecasts with their eventual outcomes. Well-
established alternatives are the mean absolute scaled error
(MASE) and the mean squared error. These all summa-
rize performance in ways that disregard the direction of
over- or under- prediction; a measure that does place em-
phasis on this is the mean signed difference.
Where a prediction model is to be fitted using a selected
performance measure, in the sense that the least squares
approach is related to the mean squared error, the equiv-
alent for mean absolute error is least absolute deviations.

2 See also
• Least absolute deviations

• Mean absolute percentage error

1
2 4 TEXT AND IMAGE SOURCES, CONTRIBUTORS, AND LICENSES

4 Text and image sources, contributors, and licenses

4.1 Text
• Mean absolute error Source: https://en.wikipedia.org/wiki/Mean_absolute_error?oldid=753845027 Contributors: Fnielsen, Michael
Hardy, Ec5618, Rich Farmbrough, Keenan Pepper, Jonathan de Boyne Pollard, Dbfirs, SmackBot, Mjg3456789, Cydebot, Betacom-
mandBot, Phopon, David Eppstein, Insanik, Melcombe, Addbot, Fgnievinski, AndersBot, Yobot, Erik9bot, Slawekb, Backwaterbassn,
Marcocapelle, Jarrod.etc2450 and Anonymous: 10

4.2 Images
• File:Question_book-new.svg Source: https://upload.wikimedia.org/wikipedia/en/9/99/Question_book-new.svg License: Cc-by-sa-3.0
Contributors:
Created from scratch in Adobe Illustrator. Based on Image:Question book.png created by User:Equazcion Original artist:
Tkgd2007
• File:Text_document_with_red_question_mark.svg Source: https://upload.wikimedia.org/wikipedia/commons/a/a4/Text_document_
with_red_question_mark.svg License: Public domain Contributors: Created by bdesham with Inkscape; based upon Text-x-generic.svg
from the Tango project. Original artist: Benjamin D. Esham (bdesham)

4.3 Content license

• Creative Commons Attribution-Share Alike 3.0