Agglomerative Independent Variable Group Analysis
Antti Honkela, Jeremias Seppä and Esa Alhoniemi
In: 15th European Symposium on Artificial Neural Networks (ESANN 2007), 25-27 Apr 2007, Bruges, Belgium.
Independent Variable Group Analysis (IVGA) is a principle for
grouping dependent variables together while keeping mutually
independent or weakly dependent variables in separate groups.
In this paper an
agglomerative method for learning a hierarchy of IVGA groupings is
presented. The method resembles hierarchical clustering, but the
distance measure is based on a model-based approximation of mutual
between groups of variables. The approach also allows
determining optimal cutoff points for the hierarchy. The method is
demonstrated to find sensible groupings of variables that ease
construction of a predictive model.