Loading...
Please wait, while we are loading the content...
Similar Documents
Discriminative log-linear grammars with latent variables (2008)
| Content Provider | CiteSeerX |
|---|---|
| Author | Klein, Dan Petrov, Slav |
| Description | In Proceedings of NIPS 20 |
| Abstract | We demonstrate that log-linear grammars with latent variables can be practically trained using discriminative methods. Central to efficient discriminative training is a hierarchical pruning procedure which allows feature expectations to be efficiently approximated in a gradient-based procedure. We compare L1 and L2 regularization and show that L1 regularization is superior, requiring fewer iterations to converge, and yielding sparser solutions. On full-scale treebank parsing experiments, the discriminative latent models outperform both the comparable generative latent models as well as the discriminative non-latent baselines. 1 |
| File Format | |
| Publisher Date | 2008-01-01 |
| Access Restriction | Open |
| Subject Keyword | Discriminative Non-latent Baseline Comparable Generative Latent Model Discriminative Training Log-linear Grammar Discriminative Log-linear Grammar Discriminative Method Full-scale Treebank Latent Variable L2 Regularization Discriminative Latent Model Gradient-based Procedure Sparser Solution Hierarchical Pruning Procedure Feature Expectation L1 Regularization |
| Content Type | Text |
| Resource Type | Proceeding |
