|WikiProject Futures studies||(Rated C-class, High-importance)|
|WikiProject Statistics||(Rated C-class, Mid-importance)|
Concern with classification of methods
Regression analysis, ARMA, and ARIMA are not causal methods (they do not "identify the underlying factors that might influence the variable"). They should be moved to time-series methods, or Causal and econometric methods should be separated. I think this page needs some work. I'm not fluent enough in this area to make edits, though. Clebio 16:14, 6 November 2006 (UTC)
- I agree that ARMA, ARIMA are not causal methods, although I had thought multiple regression was one. There are other forecasting methods I am aware of that are not mentioned - wavelets, transfer functions, singular spectrum analysis, Fourier analysis, naive methods, and many others. 22.214.171.124 21:18, 26 April 2007 (UTC)
Proposal to merge with Predictive analytics
- Merge tag removed. Se discussion in "Predicitve analytics."CommodiCast 15:41, 15 September 2006 (UTC)
I don't usually edit pages so I don't really know wikipedia policy, but the following seems like advertising a website.
"The main source of information about forecasting on the internet is the Forecasting Principles site, forecastingprinciples.com." —The preceding unsigned comment was added by 126.96.36.199 (talk) 15:50, 1 March 2007 (UTC).
I have to agree the second paragraph on this article sounds very much like advertising. Some of the external links might be useful and should be considered to be moved to a different section possibly.
- As someone who has taken an interest in the subject, I can confirm that the comment about the Forecasting Principles site is true and fair. It is much bigger than the others. The other links are useful and should be kept. 188.8.131.52 21:24, 26 April 2007 (UTC)
I too don't usually edit pages but I know that eGain is a company, among others, that provide forecast based building control. This is advertising. — Preceding unsigned comment added by 184.108.40.206 (talk) 08:44, 3 October 2013 (UTC)
Software for forecasting
There are dozens of programs that can do forecasts. It would be nice to have a seperate article that compared their various methods used etc. Some are purpose built, others include forecasting among other statistical things. Some are free, others commercial. A few do forecasting automatically.
Off the top of my head ones I recall include Autobox, Forecast Pro, R, Gretl, Easyreg, and dozens of others. 220.127.116.11 21:36, 26 April 2007 (UTC)
It seems that people involved in forecasting sometimes use the term "econometric model" as roughly synonymous with models of causation. In economics, the term is used somewhat more broadly, I think, as the application of statistical methods to economics. Regardless of the appropriate use of the term, it makes no sense to have it nestled between regression and ARIMAX as an example of a "causal/econometric" method. Might as well have "statistics" listed. --Sjsilverman (talk) 23:27, 30 November 2010 (UTC)
There is a factual conflict between two Wikipedia articles on forecasting.
Forecasting accuracy (this article): "Forecasting accuracy, in contrary to belief, cannot be increased by the addition of experts in the subject area relevant to the phenomenon to be forecast."
Weather_forecasting#Nowcasting: "A human given the latest radar, satellite and observational data will be able to make a better analysis of the small scale features present and so will be able to make a more accurate forecast for the following few hours."
Both statements cite references. I believe that weather forecasting provides an exception to the general rule (which is thus not always true). Given that the public is exposed weather forecasts every day, I believe that this is the article to be clarified.18.104.22.168 (talk) 23:45, 17 February 2012 (UTC)
- The first (this) article is talking about adding more "experts". The second (weather forecasting) is talking about a comparison of a computer moded operating at a coarse time/space resolution (and unable to make use of everything that is known) with what can be done by an expert (or more realistically, a better computer model) who is able to make use of everthing that is known. Thus adding more information on which to base foreasts is useful, adding more experts is not (if you believe the source cited). Melcombe (talk) 12:30, 18 February 2012 (UTC)