Published December 2011
| public
Book Section - Chapter
Learning Mixtures of Tree Graphical Models
Abstract
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.
Additional Information
The first author is supported in part by the NSF Award CCF-1219234, AFOSR Award FA9550-10-1-0310, ARO Award W911NF-12-1-0404, and setup funds at UCI. The third author is supported by the NSF Award 1028394 and AFOSR Award FA9550-10-1-0310.Additional details
- Eprint ID
- 118591
- Resolver ID
- CaltechAUTHORS:20221222-212034677
- CCF-1219234
- NSF
- FA9550-10-1-0310
- Air Force Office of Scientific Research (AFOSR)
- W911NF-12-1-0404
- Army Research Office (ARO)
- University of California, Irvine
- OIA-1028394
- NSF
- Created
-
2022-12-23Created from EPrint's datestamp field
- Updated
-
2022-12-23Created from EPrint's last_modified field