训练高斯混合模型的程序，尽管此类代码较多，但本程序经过笔者改写优化后，很大程度上避免了普通方法中局部最优的问题。-Gaussian mixture model training procedures, although the code more, but the procedure after the author rewrite optimization, largely avoiding the ordinary method of optimal local issues.

采用期望最大算法最优化初始高斯混合模型的程序，笔者借用ubm自适应方法，很好的解决了模型不收敛的问题。-expectations largest algorithm using optimization initial Gaussian mixture model procedures, the author borrowed ubm adaptive method, a good model is not the solution convergence problem.

计算高斯混合模型先验概率和后验概率的程序，采用大矩阵运算，大大提高了运行速度。-Gaussian mixture model calculated a priori probability and the probability of post-mortem procedures, using a large matrix computation, greatly improved speed.

高斯混合模型[Gaussian mixture model,简称GMM]是单一高斯机率密度函数的延伸,由於GMM 能够平滑地近似任意形状的密度分布,因此近年来常被用在语音与语者辨识,得到不错的效果。 -Gaussian mixture model [Gaussian mixture model, referred to as GMM] are single-Gaussian probability density function of the extension.GMM can approxi