Re: time series prediction with octave

[fotios]

On Thu, 2011-10-13 at 16:57 +0300, George Kousiouris wrote:
> 
> Hmmm, non-chaotic are the ones that are strictly periodical? 
> 
> I would say that there could be some periodicity, but definitely
> non-deterministic. I have in mind time series for example from website
> workload (user requests). This could have periodic features (like
> daytime/nighttime traffic) but can also have random events that
> disturb this behaviour (for example breaking news etc.)   
> 
> 
> On 10/13/2011 4:35 PM, Doug Stewart wrote: 
> > 
> > 
> > On Thu, Oct 13, 2011 at 4:06 AM, George Kousiouris
> > <address@hidden> wrote:
> >         
> >         Hi all,
> >         
> >         I was wandering if there is any package/function in octave
> >         that performs time series prediction.
> >         
> >         BR,
> >         George
> >         
> >         -- 
> >         
> > 
> > 
> > 
> > 
> > That is a big subject. Are you looking at a chaotic or non chaotic
> > series?
> > 
> > 
> > Doug
> >  
> > 
> > 
> > -- 
> > DAS
> > 
> > 
> > https://linuxcounter.net/user/206392.html
> > 
> > _______________________________________________
> > Help-octave mailing list
> > address@hidden
> > https://mailman.cae.wisc.edu/listinfo/help-octave
> 
> 
> -- 
> 
> ---------------------------
> 
> George Kousiouris
> Electrical and Computer Engineer
> Division of Communications,
> Electronics and Information Engineering
> School of Electrical and Computer Engineering
> Tel: +30 210 772 2546
> Mobile: +30 6939354121
> Fax: +30 210 772 2569
> Email: address@hidden
> Site: http://users.ntua.gr/gkousiou/
> 
> National Technical University of Athens
> 9 Heroon Polytechniou str., 157 73 Zografou, Athens, Greece
> _______________________________________________
> Help-octave mailing list
> address@hidden
> https://mailman.cae.wisc.edu/listinfo/help-octave

Nope, non-chaotic are those with n-cycles (period n) plus those that are
stochastic - if i recall right. I think implementing yourself Takens
time-embedding is both easy and feasible in your case - can be used for
series that "look" stochastic due to their dependence on many variables
and the method constructs/isolates those state variables that contribute
more. A very nice explanation - simplified but complete in greek - is
provided by Tasos Bountis at his book "O thavmastos kosmos ton
fractal" (the amazing fractal world). Hope it helps a bit.

/Fotis