Hello all
I'm commencing work on a univariate forecast using the ARIMA method and
have been trying to decide how many time periods ahead we can forecast
given the amount of data points we have initially.
For example, four years of weekly data gives us around 200 data points.
If we hold back the last 50 points for validation, this gives us 150
data points to develop a forecast model. How many weeks beyond the 200
data points could be predicted accurately? Could we predict for a whole
year (50 points) beyond our 200 data points?
I assume that the accuracy of the model and prediction interval would
influence our decision on the maximum length of forecast.
I've done a allstat archive search and a www lit search to find some
"rules of thumb" but to no avail.
Any advice would be much appreciated.
Many thanks
Chris
Chris Chumun
Senior Marketing Science Executive
Millward Brown
Olympus Avenue, Tachbrook Park, Warwick, CV34 6RJ
t: +44 (0) 1926 826335
f: +44 (0) 1926 833600
www.millwardbrown.com
[log in to unmask]
====================================================
This email is confidential and intended solely for the use of the
individual or organisation to whom it is addressed. Any opinions or
advice presented are solely those of the author and do not necessarily
represent those of the Millward Brown Group of Companies. If you are
not the intended recipient of this email, you should not copy, modify,
distribute or take any action in reliance on it. If you have received
this email in error please notify the sender and delete this email
from your system. Although this email has been checked for viruses
and other defects, no responsibility can be accepted for any loss or
damage arising from its receipt or use.
====================================================
|