Re: ABOUT BOX JENKINS

On Sep 11, 8:32 pm, "Rajiv Singh" <Rajiv.Si...@xxxxxxxxxxxxx> wrote:
Please see function BJ in System Identification Toolbox.

Rajiv<arda2000i...@xxxxxxxxx> wrote in message

news:1170918363.731411.44960@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx



Hi everyone. I am trying to applybox jenkinsmethod in matlab.But i
found something in the GARCH toolbox.Can i do this analyses in matlab?
IF yes how? IF no can you recommend me an alternative program.- Hide quoted text -

- Show quoted text -

Box-Jenkins is definitely the preferred approach...

Just make sure that ..

1. There are no Pulses , Level Shifts , Seasonal Pulses or Local Time
Trends in the residuals. If there are then simply add thet structure
to your Box-Jenkins Model.

2. That you verify(test) the assumption that the paramaters are
invariant over time. You can do this simply by employing a CHOW TEST
for constancy of parameters.

3. That the variance of the residuals is constant over time and is not
dependent on the level of the series .

hope this helps ..

dave reilly
Automatic Forecasting Systems
http://www.autobox.com




.

Relevant Pages

  • Re: Portmanteau test of residuals
    ... You also have to make sure that the residuals that you are analyzing ... are free of Pulses, Level Shifts, Seasonal Pulses and Local Time Trends ...
    (sci.stat.math)
  • Re: Portmanteau test of residuals
    ... You also have to make sure that the residuals that you are analyzing ... are free of Pulses, Level Shifts, Seasonal Pulses and Local Time Trends ...
    (sci.stat.math)