Displaying 2 results from an estimated 2 matches for "varepsilon_".
2010 Oct 04
0
spatial interaction (gravity) model as Poisson regression
...here are several mentions of this question,
but I have yet to find anything concrete that I can wrap my head
around... apologies if I have missed something.
Basically, the conventional origin constrained model would look
something like this:
T_{ij} = exp(\delta_{i} + \log{A_{j}} - \beta D_{ij}) ~ \varepsilon_{ij}
where \delta_{i} is a constant parameter speci?c to the ith zone,
A_{j} is the attractiveness of the jth location, and D_{ij} is the
distance between i and j. Note that \varepsilon_{ij} is just the
multiplicative error term of the ?ow from i to j, and \beta is the
distance decay parameter.
Sim...
2009 Feb 08
0
Initial values of the parameters of a garch-Model
...brary.
After fitting a ar(3)-garch(1,1)-model
> model<-garchFit(~arma(3,0)+garch(1,1), analyse)
I'm saving the results via
> result<-model at fit$se.coef
I'm wondering how the first variance
> result at h.t[1]
is computed, i.e. how the initial parameters for \varepsilon_{t-1} and
\sigma_{t-1}^2 have been chosen so that the first variance is
[1] 0.003381579
and why they have been chosen this way.
Even after snooping the function-code I cannot find it.
Any help would be appreciated.
Thanks,
Thomas
Here is the data used:
dput(analyse)
structure(list(V1 = c...