论文部分内容阅读
在传统的隐马尔可夫模型中，模型在某状态停留一定时间的概率随着时间的增长呈指数下降的趋势．文中使用依赖于时间的状态转移概率对状态停留时间予以刻画．首先，在采用相同特征矢量下进行了修改后的隐马尔可夫模型和传统隐马尔可夫模型的比较和分析．其次，对不同特征矢量的组合进行了对比实验．另外，在进行不同参数的组合时，文中考虑了不同特征参数及其维数对观察矢量概率输出的影响．
In the traditional Hidden Markov Model, the probability that a model stays in a certain state for a certain period of time tends to decrease exponentially with time. In this paper, state-dependent dwell time is described by using time-dependent state transition probability. First, the comparison and analysis between the modified Hidden Markov Model and the traditional Hidden Markov Model are carried out under the same feature vector. Secondly, the contrast experiments of different feature vectors are carried out. In addition, when the combination of different parameters is carried out, the effect of different feature parameters and their dimensions on the output of observation vector probability is considered.