Incremental Bayesian networks for structure prediction

12 years 10 months ago
Incremental Bayesian networks for structure prediction
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 e an appropriate abstract model of structure prediction tasks.
Ivan Titov, James Henderson
Added 17 Nov 2009
Updated 17 Nov 2009
Type Conference
Year 2007
Where ICML
Authors Ivan Titov, James Henderson
Comments (0)