Skip to content

[source]

Mean Absolute Error#

A scale-dependent metric that measures the average absolute error between a set of predictions and their ground-truth labels. One of the nice properties of MAE is that it has the same units of measurement as the labels being estimated.

MAE=1ni=1n|YiYi^|

Note

In order to maintain the convention of maximizing validation scores, this metric outputs the negative of the original score.

Estimator Compatibility: Regressor

Output Range: -∞ to 0

Parameters#

This metric does not have any parameters.

Example#

use Rubix\ML\CrossValidation\Metrics\MeanAbsoluteError;

$metric = new MeanAbsoluteError();

Last update: 2021-01-26