arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 0909.1308cs.LGcs.CL
收藏

稀疏条件随机场在监督序列标号中的有效学习

Efficient Learning of Sparse Conditional Random Fields for Supervised Sequence Labelling

Telecom ParisTech(巴黎高等电信学院);LTCI CNRS(法国国家科学研究中心LTCI实验室);Université Paris-Sud 11(巴黎第十一大学);LIMSI CNRS(法国国家科学研究中心LIMSI实验室)

Nataliya Sokolovska, Thomas Lavergne, Olivier Cappé, François Yvon

摘要:

条件随机场(CRF)是一种流行的和有效的监督序列标记的方法。CRF可以处理大的描述空间,并且可以在标签之间集成某种形式的结构依赖性。在这方面的贡献,我们解决的问题,有效的特征选择CRF的基础上施加稀疏通过L1罚款。我们首先展示了如何利用参数集的稀疏性来显着加快训练和标记。然后,我们引入坐标下降的参数更新计划CRF与L1正则化。最后,我们提供了一些实证比较所提出的方法与国家的最先进的CRF培训策略。特别是,它表明,所提出的方法是能够利用的稀疏性,以加快处理,因此有可能处理更大的维度模型。

英文摘要:

Conditional Random Fields (CRFs) constitute a popular and efficient approach for supervised sequence labelling. CRFs can cope with large description spaces and can integrate some form of structural dependency between labels. In this contribution, we address the issue of efficient feature selection for CRFs based on imposing sparsity through an L1 penalty. We first show how sparsity of the parameter set can be exploited to significantly speed up training and labelling. We then introduce coordinate descent parameter update schemes for CRFs with L1 regularization. We finally provide some empirical comparisons of the proposed approach with state-of-the-art CRF training strategies. In particular, it is shown that the proposed approach is able to take profit of the sparsity to speed up processing and hence potentially handle larger dimensional models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑