Correlated non-parametric latent feature models
Finale Doshi and Zoubin Ghahramani
In: UAI 2009, 18-21 JUN 2009, Montreal, Quebec.
We are often interested in explaining data
through a set of hidden factors or features.
When the number of hidden features is un-
known, the Indian Buffet Process (IBP) is
a nonparametric latent feature model that
does not bound the number of active features
in dataset. However, the IBP assumes that
all latent features are uncorrelated, making
it inadequate for many realworld problems.
We introduce a framework for correlated non-
parametric feature models, generalising the
IBP. We use this framework to generate sev-
eral specific models and demonstrate appli-
cations on realworld datasets.