Date:         Fri, 29 Mar 1996 09:31:50 EST
Reply-To:     "Name: Matthias Kehder" <mkehder@EPSILON.COM>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         "Name: Matthias Kehder" <mkehder@EPSILON.COM>
Subject:      Re[2]: WHAT'S THE BEST FORECASTING SOFTWARE?
Comments: To: Kenneth <bz7210532@GAMMA.NTU.AC.SG>

SPSS is very good for forecasting.  It does have limitiations however.  If you
would like to do within case transformations, you are likely to pull out some of
your hair, trying to figure out their vector calculations.  For auto-regressive,
moving average, exponential, and ARIMA models, SPSS works great.  Spectral
analysis is a little more complicated.  It is easy to do the actual spectral,
i.e. spectral density and periodograms (spelling?), but to do actual predictions
requires some in depth knowledge of spectral analysis (then again, so does any
other forecasting).  If you have a somewhat simple time-series SPSS is the way
to go.  If you want to do some more crazy stuff, be prepared to use some other
tools in combination with SPSS to get your data to look the way you want it to
look ... forecast pro is a waste of time.  Unless you have a picture perfect
problem (or are an MBA), it won't do what you want it to do ...
 
Matthias
 
EPSILON
 
______________________________ Reply Separator _________________________________
Subject: Re: WHAT'S THE BEST FORECASTING SOFTWARE?
Author:  Kenneth <bz7210532@GAMMA.NTU.AC.SG> at internet
Date:    3/29/96 8:59 AM
 
 
Dear friends,
 
        Thanks for the help you've given me. Many of you have recommended
various types of software that will be useful for statistical forecasting e.g.
SAS/ETS, Forecast Pro...etc.
 
        BTW, somebody else mentioned SPSS (esp. SPSS Trends). Does anyone know
whether it's a viable option? What're its strengths or if there are any
weaknesses with SPSS, could someone like to share?
 
        Thanks for the help, everybody!
 
Have a nice coming Easter!
 
Kenneth.