MDS on longitudinal data

["Bhattacharyya, Arunava, ALABS" <arunb@att.com>]

This is not a S+ question, but I fee this the right group to seek help form.

 

I am trying to find good references - both theory and applications - on how to use MDS on longitudinal data where N is slightly bigger than T. I a have data on 30 units over 24 time points. New variables are entering into the dataset (which makes it richer over time) over time as technology changes. I like to see how cluster/groups membership of these 30 units are changing over time. I am familiar with MDS on cross-sectional data, but in this analysis I like to measure change in group composition over time.

 

Any help on this is high;y appreciated. As I mentioned I like to get references on theory and applied work in this area. 

 

Thanks!

 

Arun Bhattacharyya

AT&T Labs

Florham Park.

NJ