- MIRA (Margin Infused Relaxed Algorithm )一种超保守在线算法, 在分类、排序、预测等应用领域取得不错成绩
理论
Koby Crammer.
2004. Online Learning of Com-plex Categorial Problems. Hebrew Univeristy of Jerusalem, PhD Thesis
2005. Ryan McDonald, and Fernando Pereira.Scalable large-margin online learning for structured classification.In NIPS Workshop on Learning With Structured Outputs
2006.Dekel O, Keshet J, Shalev-Shwartz S, SingerY. Online passive-aggressive algorithms.Journal of Machine Learning Research, 7: 551¡585
Ryan McDonald
2005.Crammer K, Pereira F. Online large-margin training of dependency parsers. In: Proceedings of the 43rd Annual Meeting on the Association for Computational Lin-guistics. Ann Arbor, USA: ACL, . 91¡98
2005. Femando Pereira, Kiril Ribarow, and Jan Hajic. Non-projective dependency parsing using spanning tree algorithms. In Proceedings of HLT/EMNLP, pages 523–530.
2006.Discriminative Training and Spanning Tree Algorithms for Dependency Parsing. University of Pennsylvania, PhD Thesis.
使用
2009.An Error-Driven Word-Character Hybrid Model for Joint Chinese Word Segmentation and POS Tagging
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CRF(Conditional random fields)条件随机场,2001开始用于序列标注问题
理论
John Lafferty
2001. Andrew McCallum, and Fernando Pereira. Conditional random fields: Probabilistic models for segmenting and labeling sequence data. InProceedings of ICML, pages 282–289
使用
2012. Fast Online Training with Frequency-Adaptive Learning Rates for Chinese Word Segmentation and New
未完待续……