Mean absolute error

Mean Absolute Error (MAE) is a common metric used to evaluate the performance of regression models. It measures the average magnitude of the errors in a set of predictions, without considering their direction (i.e., it takes the absolute values of the errors).