Model Initialization 

Use

Before the system calculates forecast values, it executes a model initialization for the following forecast models:

●     First-order exponential smoothing

●     Second-order exponential smoothing

●     Seasonal trend model

Prerequisites

●     You have specified at least 3 in the Historical Periods field on the General tab page in the forecast profile.

You get to the forecast profile on the SAP Easy Access screen under Advanced Planning and Optimization → Service Parts Planning (SPP) → Planning → Forecasting → Forecast Profile.

●     You have executed the IMG activity Initialize Base Values and defined weighting factors for a trip length of three.

For more information, see the Implementation Guide (IMG) for Advanced Planning and Optimization under Supply Chain Planning → Service Parts Planning (SPP) → Forecasting → Initialize Base Values.

Features

The system either performs model initialization using tripping or it uses the first three values of the demand history to calculate the base value.

You can define the number of periods that you want to consider in the forecast profile on the General tab page in the Historical Periods parameter.

Calculate Base Value

The system proceeds as follows to calculate the base value:

It weighs the first three periods of the considered demand history with the weighting factors that you specified in the Initialize Base Values IMG activity. In doing so, the system uses the weighting factors that you assigned to the trip length of three. On the basis of this base value, the system calculates the forecast from the remaining periods of the demand history.   

You have specified 36 as the number of historical periods to be considered.

Your entries in the IMG activity Initialize Base Values include the following:

Trip Length                Weighting Factor

3                                 30

3                                 30

3                                 40

The first three periods of the considered demand history have the following historical data:

Period 1 (current period – 36 + 1):    21 pieces

Period 2 (current period – 35 + 1):    15 pieces

Period 3 (current period – 34 + 1):    16 pieces

In this case, the base value for the model initialization is calculated as follows:

Base value = 30% of historical value of period 1 + 30% of historical value of period 2 + 40% of historical value of period 3

= 6,3 + 4,5 + 6,4 = 17,2

On the basis of this base value, the system calculates the forecast from the remaining 33 periods of the demand history.

Trend Value

For the seasonal trend model, the system sets the trend value to zero.