Return to search

Methods and Experiments With Bounded Tree-width Markov Networks

Markov trees generalize naturally to bounded tree-width Markov networks, onwhich exact computations can still be done efficiently. However, learning themaximum likelihood Markov network with tree-width greater than 1 is NP-hard, sowe discuss a few algorithms for approximating the optimal Markov network. Wepresent a set of methods for training a density estimator. Each method isspecified by three arguments: tree-width, model scoring metric (maximumlikelihood or minimum description length), and model representation (using onejoint distribution or several class-conditional distributions). On thesemethods, we give empirical results on density estimation and classificationtasks and explore the implications of these arguments.

Identiferoai:union.ndltd.org:MIT/oai:dspace.mit.edu:1721.1/30511
Date30 December 2004
CreatorsLiang, Percy, Srebro, Nathan
Source SetsM.I.T. Theses and Dissertation
Languageen_US
Detected LanguageEnglish
Format10 p., 10714507 bytes, 473643 bytes, application/postscript, application/pdf
RelationMassachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory

Page generated in 0.0019 seconds