Re: How do I include causals into an existing forecast model?
- From: dave@xxxxxxxxxxx
- Date: Thu, 29 May 2008 05:25:46 -0700 (PDT)
On May 27, 7:15 pm, Kiran <kiranga...@xxxxxxxxx> wrote:
Hello,
I am trying to forecast retail sales in an excel based model. I have
built several ARIMA, Smoothing and Winters models in excel to include
trend & seasonality. However I also have information on the causals
such as Advertisements, which I want to include in the model. How do I
include the causals into the model? Is there a way we can add
regression to ARIMA (in excel?). Or may be there is another way to do
it? (Remove Ad factor before forecasting and then add it back??). Has
anyone done this before? Thanks in advance!
-Kiran
Kiran,
ARIMA is rear-window driving or rear-window forecasting
TRANSFER FUNCTIONS incorporate both ARIMA and Regressor Variables.
Intervention DEtection can be useful in both cases to uncover
unspecified DETERMINISTIC VARIABLES
Go to http://www.autobox.com and SEARCH for these tertms ... Enroll at
AFS university to lean more about the subject of forecasting with
causals ... when you have time series data .
Holt-Winters models are REAR-WINDOW versions of ARIMA Models and while
simple to use and understand are much to SIMPLE to adequately separate
signal from noise.
You might also downlaod a piece of FREEWARE called appropriately
FREEFORE from http://www.autobox.com/freef.exe
and if you would like some personal help .... feel free to call
Dave R
215-675-0652 in the US
.
- References:
- Prev by Date: How do I include causals into an existing forecast model?
- Previous by thread: How do I include causals into an existing forecast model?
- Index(es):
Relevant Pages
|
Loading
