The mean absolute error (MAE) measures how close the model's predictions are to the actual target values. It is the sum of the absolute value of the difference of each instance prediction and the actual value. For classification, the 0/1-error is used. $$\mathrm{MAE} = \frac{1}{n}\sum_{i=1}^n \left| f_i-y_i\right| =\frac{1}{n}\sum_{i=1}^n \left| e_i \right|.$$ See: http://en.wikipedia.org/wiki/Mean_absolute_error
| Minimum value | 0 |
| Maximum value | 1 |
| Unit | |
| Optimization | Lower is better |