Using the residuals, we can measure the error from the predicted and actual values based upon three popular accuracy measures:
Mean absolute error (MAE): This measure takes the mean of the absolute values of all of the errors (residuals)
Root-mean-squared error (RMSE): The root mean square error measures the error by first taking the mean of all of the squared errors, and then takes the square root of the mean, in order to revert back to the original scale. This is a standard statistical method of measuring errors.
Note
Both MAE and RMSE are scale-dependent measures, which means that that they can be used to compare problems with similar scales. When comparing accuracy among models with different scales, other scale-independent measures such as MAPE should be used.
Mean percentage error (MAPE): This is the absolute difference between the actual and forecasted value, expressed as a percentage of the actual value. This is intuitively easy to understand and is a very popular measure...
