Informatics Report Series


Report   

EDI-INF-RR-1013


Related Pages

Report (by Number) Index
Report (by Date) Index
Author Index
Institute Index

Home
Title:Incremental Bayesian Networks for Structure Prediction
Authors: Ivan Titov ; James Henderson
Date:Jun 2007
Publication Title:In Proc. 24th International Conference on Machine Learning (ICML 2007)
Publication Type:Conference Paper Publication Status:Published
Page Nos:887-894
Abstract:
We propose a class of graphical models appropriate for structure prediction problems where the model structure is a function of the output structure. Incremental Sigmoid Belief Networks (ISBNs) avoid the need to sum over the possible model structures by using directed arcs and incrementally specifying the model structure. Exact inference in such directed models is not tractable, but we derive two efficient approximations based on mean field methods, which prove effective in artificial experiments. We then demonstrate their effectiveness on a benchmark natural language parsing task, where they achieve state-of-the-art accuracy. Also, the model which is a closer approximation to an ISBN has better parsing accuracy, suggesting that ISBNs are an appropriate abstract model of structure prediction tasks.
Links To Paper
1st Link
2nd Link
Bibtex format
@InProceedings{EDI-INF-RR-1013,
author = { Ivan Titov and James Henderson },
title = {Incremental Bayesian Networks for Structure Prediction},
book title = {In Proc. 24th International Conference on Machine Learning (ICML 2007)},
year = 2007,
month = {Jun},
pages = {887-894},
url = {http://homepages.inf.ed.ac.uk/jhender6/papers/titov_icml07.pdf},
}


Home : Publications : Report 

Please mail <reports@inf.ed.ac.uk> with any changes or corrections.
Unless explicitly stated otherwise, all material is copyright The University of Edinburgh