on 08-04-2009 1:05 PM
hi sir
good evening
can any body tell the methods ( univerate , MLR, and Composite) we are using in dp , which is the best method for the business?
and which senerio which method would use?? can anyone help me out??
2. what is the different between MLR method and Casual Analysis?
3. In univerate and MLR can available the erros of measurement. why can't those kind of erros not availble for Composite??
4.what is the mean of Statistical forecasting. is that same as Univerate??
5. what is the collebarative forecasting and consensed based forecasting???
Hi Dallyanusha,
I have attached a link which contains the answers for all your queries.
http://help.sap.com/saphelp_scm70/helpdata/EN/19/98ad1765354d7ba54b1eb164c377e0/frameset.htm
Please confirm
Regards,
R. Senthil Mareeswaran.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
HI,
This will be a broad question and difficult answer, you are expecting the entire DP in one question. i will try to summerise the same,
1. we have forecast profiles - Univariate, Causal and composite
2. what profile you need to use is entirely depens on your data patterns, let say it has promation activities in past then we will use casual analysis model.
3. we have techniques in each profile like univarite - LR, SLR, Holts winters like and similarly you have MLR for causal analysis
4. there will error measures for univariate and causual but not for composire becuase composite is really not a techniq, it just an weigheted avg of both the models. if you want to use both univariate and MLR for the same data then run both of them take the avg of it like, the composite wil help you to take the avg or % of wighing.
Yarala
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
9 | |
4 | |
3 | |
2 | |
2 | |
1 | |
1 | |
1 | |
1 | |
1 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.