Exponential Smoothing

Applies the Exponential Smoothing method to the series' values. It is included in Smoothing and Forecasting groups.

The Exponential Smoothing method is a technique of averaging time series values. In contrast to moving average method, all observations of source time series take part in calculating exponential mean, but these observations have different weight coefficients. The time of the time series value plays the key role for the exponential smoothing. Exponentially decreasing weights are assigned to older observations; all previous series observations are taken into account.

To apply the method

After applying the method in the workbook, a calculated series containing calculation results is created on the base of each selected series with the name of the following type:

• On executing the method from the Smoothing category - Exponential Smoothing(<Series_Name>). For example:

• On executing the method from the Forecasting category - Exponential Smoothing: forecast(<Series_Name>). For example:

Setting Up Calculation Parameters

To set up specific calculation parameters, use side panel tabs:

• Parameters. Enables the user to change basic calculation parameters: seasonality and growth models.

• Parameters Autofit. Enables the user to set up parameters of coefficient values autofit used on method calculation.

See also:

Working with Calculated Series | Exponential Smoothing | Modeling ContainerExponential Smoothing | IModelling.Expsmooth | IMsExponentialSmoothingTransform