SquareLoss¶

class getml.models.loss_functions.SquareLoss¶

Bases: getml.models.loss_functions._LossFunction

Square loss (aka mean squared error (MSE))

Measures the loss by calculating the average of all squared deviations of the predictions $\hat{y}$ from the observed (given) outcomes $y$ . Depending on the context this measure is also known as mean squared error (MSE) or mean squared deviation (MSD). deviation (MSD).

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

with $n$ being the number of samples, $y$ the observed outcome, and $\hat{y}$ the estimate.

Note

Recommended loss function for regression problems.