Statistical Modelling 10 (2010), 41–56
Discriminant partial least squares analysis on compositional data
Michele Gallo
Department of Social Science
University of Naples L’Orientale
Largo S. Giovanni Maggiore
I–80134 Naples
Italy
eMail: mgallo@unior.it
Abstract:
Compositional data are commonly present in many disciplines. Nevertheless,
it is often improperly incorporated into statistical modelling and a
misleading interpretation of the results is given. This paper explains
how partial least squares for discrimination is an adequate technique
for compositional data when a dimensional reduction of original
variables is needed and difining the variables that more influence
the discrimination between the observations is the goal.
Keywords:
Aitchison distance; compositional observation;
discriminant analysis; log contrast; partial least squares
Downloads:
Example data
back