1

Should be an easy question but Google failed me.

When using ANN for series forecasting one often uses may variables. For instance the number of shoppers might be determined by the previous number of shoppers and the previous weather conditions.

I have only ever seen ARIMA models used with one variable i.e. the number of shoppers.

Is this a feature of ARIMA models? do they only use one input (so to speak).

additional info: I've been using the ARIMA implementation provide by the Forecasting package for R. If is is possible, links to how to this is done with the forecasting package would be greatly appreciated.

Many Thanks

Andy T
  • 1,014
  • 3
  • 10
  • 16
  • I assume you are using the `forecast` package for R (as far as I know, there is neither a `Forecast` nor a `forecasting` package). Perhaps you could edit your question to clarify? (Sorry for nitpicking, but future readers will be grateful for precision.) – Stephan Kolassa Nov 21 '14 at 13:41
  • This may be helpful for how to do this in R: http://stats.stackexchange.com/questions/122803/arima-time-series-forecast-auto-arima-with-multiple-exogeneous-variables-in-r/122806#122806 – Stephan Kolassa Nov 21 '14 at 13:41

1 Answers1

2

Should be an easy question but Google failed me.

This should totally be an easy question, you only need the technical name for what you need: the "inputs" are covariates, and in the context of time series they are usually called predictors I think.

For example, Hyndman explains how to run ARIMA with covariates and he even includes a comment about R. You can find more material if you look up "time series covariates", "arima covariates", "arima covariates in r", etc, in Google.

Enjoy your time series!

mugen
  • 1,319
  • 1
  • 15
  • 24