一种基于梯度的HMM参数重估方法

对于隐Markov模型（HMM），经典的参数重估方法是Eaum-Welch算法，该算法基于最大似然准则，具有快速收敛和保证似然度单调增的优点，但是对于其他的训练准则，则不存在这样的算法，由于目标函数的复杂性，在考虑采用梯度方法时，必须先解决如何求取梯度的问题，为此，提出一种求取梯度的实现方法，结果表明，使用该方法所取得的模型与用Baum-Welch算法所得的模型性能相当，而前者适用于基于各种准则的训练方法。

A Gradient Based Estimation Method for HMMs

Baum Welch algorithm is a classical estimation method for HMMs, which is based on the maximum likelihood(ML) criterion. For other criteria, such as maximum mutual information (MMI) criterion, such an algorithm does not exist. In this case, a gradient based method is considered. With the complexity of objective function, the computation of the gradients has to be solved before it can be applied to this problem. This paper proposed an implementation method of the gradient based method. The experimental results indicate that this method produces comparable results to Baum Welch algorithm.