Describe two ways in which forecasting accuracy can be evaluated.© BrainMass Inc. brainmass.com December 24, 2021, 11:47 pm ad1c9bdddf
SOLUTION This solution is FREE courtesy of BrainMass!
Way of Evaluating Forecasting Accuracy
Scaled Error Method: It is an important method for evaluating forecast accuracy across series of different scales. It provides scaling of error on the basis of training MAE (mean absolute error) from a simple forecast method. Following is the formula to identify scaled error for a non-seasonal time series:
(See attachment for formula)
Here, qj is the value of scaled error for evaluating forecast accuracy, which is dependent on scale of data. If the value of scaled error (qj) is less than one, the forecast vales are better than average naïve forecasts. Similarly, if it is greater than one, the forecasts are worse than average forecasts (Lawrence & Klimberg, 2011).
Training and Test Sets: It is an important method for evaluating forecast accuracy using genuine forecasts. It provides that forecast accuracy can only be determined by considering how well a model performs on new data that were not used in fitting model. While choosing, it is common to use a portion of available data for fitting and use rest of data for testing the model. After this phase, testing data can be used to measure how well a model is likely to forecast on new data (Ord & Fildes, 2012). Size of test set would bet approx 20% of the total sample, although it may vary from one forecast to another based on size of forecast.
Lawrence, K.D. & Klimberg, R. (2011) Advances in Business and Management Forecasting Volume 8. UK: Emerald Group Publishing.
Ord, K. & Fildes, R. (2012) Principles of Business Forecasting. USA: Cengage Learning.