Statistical Parameters

[Former Member]

Hi,

I want to jnow the recommended values for beta and gamma.

Beta and gamma are predefined as 0.3. What are the recommended limits that are recommended? What would be the implications?

Thanks in advance

[Former Member · Accepted solution]

Hi,

Basically Alpha, Beta & gama are the smothing factors. The values of these factors to smooth History. Alpha is for basic value, Beta is for Trend value & Gama is related to Seasonal Index. There is one factor called sigma which is realted to Forecast Error.

The SAP Standard says 0.3 for all factor except Sigma. Sigma would be 1.25.

You can chge these factor & see the difference all the values like Basic, Trend & Seasonal Index. You can chnage it on Trail & Error method. You need to compaire all the forecasting data with previous run. So you can find the difference & will find out the Impact on History & forecast figure easily.

for more information you can refer followinf link

http://help.sap.com/saphelp_scm50/helpdata/en/ac/216b71337b11d398290000e8a49608/frameset.htm

Regards

Sujay

[Former Member · Accepted solution]

Hi,

As I know logical values for alpha are between 0.1 and 0.5.

What range is logical for beta and gamma?

[Former Member · Accepted solution]

Hi,

SAP Standard is also same 0.3 for beta & Gama...Range is in between 0.1 to 0.9...You can try within this range..

Regards

Sujay

[Former Member · Accepted solution]

Hi,

Values can range from .1 to .9.

You can test with various combination of beta & gamma values and check error parameters for forecast accuracy.

Based on your test results you will able to arrive at beta & gamma value. Well beta & gamma can not static for long term and you will have to revisit them from time to time and change according to the ways business statistics are changing.

Hope this helps.