0 alternative definitions
A metric used to evaluate a model's accuracy by calculating the average absolute difference between the expected and predicted values for all training examples.
