Timezone: »

Learning Mixtures of Tree Graphical Models
Anima Anandkumar · Daniel Hsu · Furong Huang · Sham M Kakade

Wed Dec 05 07:00 PM -- 12:00 AM (PST) @ Harrah’s Special Events Center 2nd Floor #None
We consider unsupervised estimation of mixtures of discrete graphical models, where the class variable is hidden and each mixture component can have a potentially different Markov graph structure and parameters over the observed variables. We propose a novel method for estimating the mixture components with provable guarantees. Our output is a tree-mixture model which serves as a good approximation to the underlying graphical model mixture. The sample and computational requirements for our method scale as $\poly(p, r)$, for an $r$-component mixture of $p$-variate graphical models, for a wide class of models which includes tree mixtures and mixtures over bounded degree graphs.

Author Information

Anima Anandkumar (Caltech)
Daniel Hsu (Columbia University)
Furong Huang (University of Maryland, College Park)
Sham M Kakade (Microsoft Research)

More from the Same Authors