Census X -12 -ARIMA ( Autoregressive Integrated Moving Average ) is a system developed by the U.S. Bureau of the Census method for seasonal adjustment of statistical time series collected. Here, as basic tools for the estimation of the most important components in economic time series, that is used by trend-cycle and seasonal components, moving averages. Prior to the estimation of these components can be carried out for the estimation of calendar effects, outliers or to reduce revisions of analysis results ARIMA modeling of time series.
On his website provides the Bureau of the Census next to a documentation and programs for Windows and Unix computer to download.
Since it is an internationally recognized method, recommended by the Federal Statistical Office for the creation of internationally comparable indices as used for example in the production index. At the Swiss Federal Statistical Office, it has replaced the parallel further applied Berlin procedure since March 2002 in the publications.