对于PCFGs和适配器语法的推理Inference for PCFGs and Adaptor Grammars |
|
课程网址: | http://videolectures.net/nipsworkshops09_johnson_ipcfgag/ |
主讲教师: | Mark Johnson |
开课单位: | 布朗大学 |
开课时间: | 2010-01-19 |
课程语种: | 英语 |
中文简介: | 本文描述了我们为适配器语法推断开发的过程。适配器语法是PCFGS的非参数扩展,可用于描述各种语音和形态语言学习任务。我们首先回顾作为适配器语法推断基础的概率上下文无关语法的MCMC采样器,然后解释如何将规则依赖于其他采样树的PCFG样本用作估计适配器语法的MCMC过程中的建议分布。最后,我们描述了几种显著加快复杂适配器语法推理的优化方法。 |
课程简介: | This talk describes the procedures we've developed for adaptor grammar inference. Adaptor grammars are a non-parametric extension to PCFGs that can be used to describe a variety of phonological and morphological language learning tasks. We start by reviewing an MCMC sampler for Probabilistic Context-Free Grammars that serves as the basis for adaptor grammar inference, and then explain how samples from a PCFG whose rules depend on the other sampled trees can be used as a proposal distribution in an MCMC procedure for estimating adaptor grammars. Finally we describe several optimizations that dramatically speed inference of complex adaptor grammars. |
关 键 词: | 适配器语法; 非参数扩展; PCFGs |
课程来源: | 视频讲座网 |
最后编审: | 2020-06-02:毛岱琦(课程编辑志愿者) |
阅读次数: | 61 |