Loss Functions

Standard

Mean Absolute Error (MAE)

Definition

$$ Err(\hat{y}_i, y_i) = \frac{1}{n} \sum_{i=1}^{n} |y_i - \hat{y}_i| $$

Mean Squared Error (MAE)

Definition

$$ Err(\hat{y}_i, y_i) = \frac{1}{n} \sum_{i=1}^{n} (y_i - \hat{y}_i)^2 $$