hello Torsten,
sorry for the delay i have just seen your replay....
thank you very much ,it helps a lot !
Is this way of calculating varience for vectors wrriten anywhere in the litrature couse when i looked for it i couldnt find it....
anyway ,thank you!
yafit
Torsten <Torsten.Hennig@umsicht.fraunhofer.de> wrote in message <f7b121aaf5e54c559a1e19ec50cb3155@s3g2000vbf.googlegroups.com>...
> On 15 Mrz., 12:38, "yafit maymon" <yafit_may...@hotmail.com> wrote:
> > hello!
> >
> > i need some help please....
> >
> > i have done some clustering to matrics with 20 random variables , each variable has 1000 observations ). i got 4 clusters
> >
> > and i need to test how good the clustering is by calculating the variance in each cluster.
> > does anyone knows how can i calculate the variance?
> >
> > i can easily calculate the variance of each coulmne in my matrix (e.g the variance of each random variable) but i want to calculate the variance of the whole cluster
> >
> > does anyone know how it can be done ?
> > thanks!!!!
>
> If
> mu_i = 1/C_i * sum_{x in C_i} x
> is the mean of cluster i, the variance within cluster i should be
> given by
> sigma_i^2 = 1/(C_i1) * sum_{x in C_i} (xmu_i)'*(xmu_i).
>
> Here C_i is the number of elements in cluster i and ' denotes
> 'transpose'.
>
> Best wishes
> Torsten.
