Another look at measures of forecast accuracy. Rob J. Hyndman, & A. B. K. International Journal of Forecasting, 22:679--688, 2006.
doi  abstract   bibtex   
We discuss and compare measures of accuracy of univariate time series forecasts. The methods used in the M-competition as well as the M3-competition, and many of the measures recommended by previous authors on this topic, are found to be degenerate in commonly occurring situations. Instead, we propose that the mean absolute scaled error become the standard measure for comparing forecast accuracy across multiple time series.
@article{rob_j._hyndman_another_2006,
	title = {Another look at measures of forecast accuracy},
	volume = {22},
	shorttitle = {Another look at measures of forecast accuracy},
	doi = {10.1016/j.ijforecast.2006.03.001},
	abstract = {We discuss and compare measures of accuracy of univariate time series forecasts. The methods used in the M-competition as
well as the M3-competition, and many of the measures recommended by previous authors on this topic, are found to be
degenerate in commonly occurring situations. Instead, we propose that the mean absolute scaled error become the standard
measure for comparing forecast accuracy across multiple time series.},
	journal = {International Journal of Forecasting},
	author = {Rob J. Hyndman, \& Anne B. Koehler},
	year = {2006},
	keywords = {Forecast accuracy, Forecast error measures, Forecast evaluation, M-competition, Mean absolute scaled error},
	pages = {679--688}
}

Downloads: 0