Root Mean Square Error
 

The accuracy of a forecasting technique can be assessed use the RMSE.
 

After the fact of forecasting, the actual (observations =n) values of the variable for the forecast period can be compared to the forecast values. A difference for each observation is computed then squared. These differences are summed after squaring and a mean is determined. Finally the square root is taken. We get a root mean square error. It is something like the average error, but more sophisticated.
 

RMSE        =