ezForecaster
  ⊲Buy ezForecaster Now!
  ⊲Why should you use ezForecaster? 
  ⊲
What is ezForecaster?
  ⊲Who is using ezForecaster?
  ⊲ezForecaster Case Studies
  ⊲
Competitive Analysis
  ⊲Forecasting in 3 Easy Steps
  ⊲Watch a Guided Tour
  ⊲FAQs

Contact Us
  ⊲Get Product Support
  ⊲Contact Sales

Behind the Scenes
  ⊲How does ezForecaster work?
  ⊲What is Time Series Forecasting?
  ⊲
21 Forecast Methods
  ⊲Automatic Best Fit selection
  ⊲White Papers
 

ezForecaster Automatically Chooses the Best Fitting Forecast - But How?

By default, ezForecaster chooses the most appropriate forecast technique by ranking three different statistical measures of Goodness of Fit: the Mean Absolute Percent Error (MAPE), the Mean Absolute Deviation (MAD) and the Root Mean Squared Error (RMSE). The Method with the lowest composite ranking is the Best Method.

Alternatively, you can let ezForecaster choose the Best Method using with the smallest MAPE, MAD or RMSE individually.

Mean Absolute Percent Error (MAPE) is calculated by averaging the percentage difference between the fitted (forecast) line and the original data. If the best fit method has a large MAPE (i.e., 40 percent or more), the forecast, for various reasons, is not a particularly good one.

MAPE = S |et/yt| * 100 / n 

where y represents the original series and e the original series minus the forecast, and n the number of observations

Mean Absolute Deviation (MAD) (also known as the Mean Absolute Error) is the Sum of Errors (the difference between the fitted line and the original data), divided by the number of data points.

MAD = S |et| / n

Root Mean Squared Error (RMSE) is the Square Root of Sum of the Squared Errors, divided by the number of data points. It is actually a measure of dispersion of the forecast from the original data.

RMSE = (S et2 / n)

 

     

 

 
Home | ezForecaster | Services | Support | About Us Copyright 1997-2016. All rights reserved.