计算机工程 ›› 2010, Vol. 36 ›› Issue (3): 197-199.doi: 10.3969/j.issn.1000-3428.2010.03.066

• 人工智能及识别技术 • 上一篇    下一篇

大间隔高斯混合模型的快速参数更新算法

黄 浩1,哈力旦2   

  1. (1. 新疆大学信息科学与工程学院,乌鲁木齐 830046;2. 新疆大学电气工程学院,乌鲁木齐 830008)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2010-02-05 发布日期:2010-02-05

Rapid Parameter Updating Algorithm for Large Margin Gaussian Mixture Model

HUANG Hao1, HAlidan2   

  1. (1. College of Information Science and Engineering, Xinjiang University, Urumuqi 830046; 2. College of Electrical Engineering, Xinjiang University, Urumuqi 830008)
  • Received:1900-01-01 Revised:1900-01-01 Online:2010-02-05 Published:2010-02-05

摘要: 针对大间隔高斯混合模型基于LBFGS参数更新算法收敛速度慢的不足,提出一种快速参数更新算法。采用构造弱意义辅助函数的方法,得到扩展Baum-Welch算法形式的快速参数更新公式。利用大词汇汉语语音库上的声调分类任务来验证训练速度与分类性能。实验结果表明快速参数更新算法只需数次迭代就能收敛至最优结果,较LBFGS优化方法在识别性能相当的情况下具有更快的训练速度。

关键词: 大间隔, 高斯混合模型, 声调识别

Abstract: To speed up training of large margin Gaussian mixture model based on LBFGS optimization routing, a rapid model parameter method is proposed. The method formulates extended Baum-Welch algorithm like updating equations by constructing weak-sense auxiliary function. The proposed algorithm is experimented on Mandarin tone recognition tasks. It is shown training can be accomplished within only several iterations, much faster than that of LBFGS optimization routine.

Key words: large margin, Gaussian mixture model, tone recognition

中图分类号: