Plan Items - Forecast Settings (cpdsp1110m000)

Field Information

Average Forecast Error

The mean forecast error that is computed during the demand forecast calculation for generating a demand plan for a plan item.

Automatic Update of Forecast Parameters

If this check box is selected, you can recalculate the forecast parameters by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

You can recalculate the following fields:

• The Type of Trend Influence field
• The Type of Seasonal Influence field
• The Seasonal Cycle Time field
• The Smoothing Factor for Demand field
• The Smoothing Factor for Trend field
• The Smoothing Factor for Season field
• The Smoothing Factor for Forecast Error field
• The Degree for Polynomial Regression field

If this check box is cleared, you must manually maintain the forecast parameters.

Channel

Specify one of the channels that you defined for the plan item in the Plan Item - Channels (cpdsp5100m000) session.

Critical Tracking Signal

If you selected the Tracking Signal for Demand Forecast check box, enter the critical value that the tracking signal for the demand forecast should not exceed.

Allowed values

Enter a value in the range 0 to 1.

If the Tracking Signal for Demand Forecast check box is selected, LN uses the tracking signal for the demand forecast during a demand calculation according to the exponential smoothing method.

If the tracking signal exceeds the critical tracking signal, the exponential smoothing method uses an increased value of the smoothing factor for demand instead of the value of the Smoothing Factor for Demand field.

The greater the value of the critical tracking signal, the less nervously the exponential smoothing method reacts to forecast errors.

A reasonable value for the critical tracking signal is between 0.5 and 8. To have an effect, the critical tracking signal must be substantially greater than the Smoothing Factor for Demand field.

Degree for Polynomial Regression

This field is only available if you selected Polynomial Regression in the Forecast Method field.

Allowed values

Enter a whole number in the range of 0 to 9.

The degree of the polynom that is used in demand forecast calculations based on the polynomial regression method.

Forecast Method

The mathematical method that LN uses to compute the demand forecast.

Each forecast method uses the historical demand data of the time period between the scenario start date and the current plan period. LN retrieves the historical demand data from the Customer Deliveries field in the Item Master Plan (cprmp2101m000) session.

Mean Absolute Deviation

The mean absolute deviation or forecast error is the average difference between the demand forecast and the actual demand.

Note: This forecast error is not the mean absolute forecast error on which the tracking signal for demand forecast is based.

Moving Average Offset

This field determines the starting point for the calculation of the moving average, and it is only available if you selected Moving Average in the Forecast Method field.

The moving average offset is the number of calendar days between:

• The starting point for the calculation of the moving average
• The forecast period

Moving Average Period

This field is only available if you selected Moving Average in the Forecast Method field.

Allowed values

Enter a value that is less than or equal to the value of the Moving Average Offset field.

The length of the period for which the moving average must be computed (in calendar days). The starting date of this period is determined by the Moving Average Offset field.

Mean Relative Deviation

The mean relative deviation or forecast error of the forecast demand against the actual demand.

Scenario

The scenario for which you set the plan item's forecast parameters in this session.

Description

The description or name of the scenario.

Plan Item

The plan item for which you maintain the forecast settings. This can be the normal plan item, or the channel-related plan item.

Note: If you specify the special demand on plan item level and not on channel level, then the special demand is input for the Special Demand field in the Item Master Plan (cprmp2101m000) session.

Seasonal Correlation Factor

The seasonal correlation factor is a relative factor, which indicates to what extent there is a seasonal cycle for a given demand pattern. The maximum value of the seasonal correlation factor is 1.0.

LN calculates the factor on the basis of the historical demand figures and the Seasonal Cycle Time field.

Seasonal Cycle Time

The cycle time of the seasonal demand pattern, expressed in calendar days.

Allowed values

Enter a value of more than 0 and less than 1000 calendar days.

If the demand for the plan item has a yearly recurring variation, the seasonal cycle time is 365 days.

If the Type of Seasonal Influence field is not Not Applicable, you cannot modify this field.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine the seasonal cycle time by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Standard Deviation

The standard deviation of the forecast error.

The standard deviation of the forecast error is the difference between the forecast demand and the actual demand, compared to the mean forecast error.

Smoothing Factor for Demand

This field is only available if you selected Exponential Smoothing in the Forecast Method field.

Allowed values

Enter a value in the range 0 to 1.

The Smoothing Factor for Demand field is a factor in the computation of the demand forecast according to the exponential smoothing method.

If the smoothing factor increases, the influence of demand data from the recent past on the demand forecast also increases. In other words, an increase of the smoothing factor will cause a stronger reaction of the demand forecast to recent demand fluctuations.

Values between 0.1 and 0.3 usually give satisfactory results.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine the smoothing factor by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Smoothing Factor for Forecast Error

This field is only available if you selected Exponential Smoothing in the Forecast Method.

Allowed values

Enter a value in the range 0 to 1.

The Smoothing Factor for Forecast Error field is a factor in the computation of the mean forecast error during the demand forecast calculation according to the exponential smoothing method.

If this factor increases, the influence of recent forecast errors on the mean forecast error also increases.

You must enter a value close to 1 (for example 0.8) in order to obtain a fast response of the mean forecast error to fluctuations in the demand.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine this smoothing factor by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Smoothing Factor for Season

The Smoothing Factor for Season field is a factor in the computation of the demand forecast according to the exponential smoothing method.

Allowed values

Enter a value in the range 0 to 1.

If the smoothing factor increases, the influence of demand data from the recent past on the seasonal influence also increases. In other words, an increase of the smoothing factor causes a stronger reaction of the demand forecast for a following seasonal cycle on recent demand fluctuations in the current seasonal cycle.

A reasonable value is something between 0.1 and 0.3.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine this smoothing factor with the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Smoothing Factor for Trend

This field is only available if you selected Exponential Smoothing in the Forecast Method field.

Allowed values

Enter a value in the range 0 to 1.

The Smoothing Factor for Trend field is a factor in the computation of the demand forecast according to the exponential smoothing method.

If the smoothing factor increases, the influence of demand data from the recent past on the trend influence also increases. In other words, an increase of the smoothing factor causes a stronger reaction of the demand forecast to recent forecast errors.

A value of 20% of the value of the Smoothing Factor for Demand field usually gives satisfactory results.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine this smoothing factor by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Type of Seasonal Influence

This field determines to what extent there is a seasonal influence on the demand forecast.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine the seasonal influence by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Type of Trend Influence

This field determines to what extent there is a trend influence on the demand forecast.

Note: If the Automatic Update of Forecast Parameters check box is selected, you can determine the trend influence by using the Generate Demand and Inventory Plan (cpdsp1210m000) session.

Tracking Signal for Demand Forecast

You can only select this check box if you selected Exponential Smoothing in the Forecast Method field.

If this check box is selected, the smoothing factor for the demand must be linked to a tracking signal during a demand forecast calculation, according to the exponential smoothing method.

The tracking signal is the ratio between the mean absolute forecast error and the mean forecast error. If the tracking signal exceeds the value of the Critical Tracking Signal field, LN automatically adjusts the smoothing factor for the demand.