I presume the reason you want to use hierarchical clustering is to
compare different slices of the tree.
Since you are using SPSS, I would suggest that you use TWOSTEP, because
it tells you the AIC/BIC for different numbers of clusters.
Art
[log in to unmask]
Social Research Consultants
University Park, MD USA
(301) 864-5570
jamie burnett wrote:
> Hi,
> I would like to run a hierachial cluster analysis on my factor scores
> (uncorrelated) to find the cluster centres of interest and then use these
> centres for my initial cluster seeds in k-means clustering. However I
> have a
> dataset with over 4000 cases. Is it possible to run hierachial cluster
> analysis in SPSS on such a large dataset or should I take a random sample
> and run it on them.
> Can anyone help?
>
> Thanks
> JB
>
> _________________________________________________________________
> Express yourself with cool new emoticons
> http://www.msn.co.uk/specials/myemo
>
>
|