public class AutoARIMAEx1 extends Object
Searches for the best fitting non-seasonal \( \text{ARIMA} \).
This example uses time series LNU03327709 from the US Department of Labor,
Bureau of Labor Statistics. It contains the unadjusted special unemployment
rate, taken monthly from January 1994 through September 2005. The values
01/2004 - 03/2005 are used by class autoARIMA
for outlier
detection and parameter estimation. Technique 1, invoked by
autoARIMA.compute(int maxlag)
is chosen to find an appropriate
\(\text{ARIMA}(p,d,0)\) model. That is, the algorithm searches for the best
fitting parameters while it is assumed that \(q = 0\). With the selected
model, forecasts are generated for the following six months and compared with
the actual values 04/2005 - 09/2005.
Constructor and Description |
---|
AutoARIMAEx1() |
Copyright © 2020 Rogue Wave Software. All rights reserved.