Tree block coordinate descent for map in graphical models
Name
Jaakkola_Tree block.pdf
Size
284.08 KB
Format
Adobe PDF
Checksum (MD5)
3ccb8a15855fdfffccc68e49a3859d83
Author(s)
Sontag, David Alexander
Jaakkola, Tommi S.
Date Issued
April 2009
Journal
Proceedings of the 12th International Conference on Artifcial Intelligence and Statistics (AISTATS) 2009
Publisher
Journal of Machine Learning Research
Citation
"Tree block coordinate descent for map in graphical models." Proceedings of the Twelfth International Conference on Artificial Intelligence and Statistics
April 16-18, 2009, Clearwater Beach, Florida USA.
Version
Author's final manuscript
Abstract
A number of linear programming relaxations have been proposed for finding most likely settings of the variables (MAP) in large probabilistic models. The relaxations are often succinctly expressed in the dual and reduce to different types of reparameterizations of the original model. The dual objectives are typically solved by performing local block coordinate descent steps. In this work, we show how to perform block coordinate descent on spanning trees of the graphical model. We also show how all of the earlier dual algorithms are related to each other, giving transformations from one type of reparameterization to another while maintaining monotonicity relative to a common objective function. Finally, we quantify when the MAP solution can and cannot be decoded directly from the dual LP relaxation.
Description
abstract URL: http://jmlr.csail.mit.edu/proceedings/papers/v5/sontag09a.html
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike 3.0
Persistent DSpace Link
DOI of Published Version
http://jmlr.csail.mit.edu/proceedings/papers/v5/sontag09a/sontag09a.pdf
