对于连续时间的层次模型Markov链,所有的转移速率都可以由最底层状态的逗留时间和击中时间分布惟一决定,因而整个Markov链的统计性质由它们的统计所决定.并给出了相应的算法和数例.
This paper proves that all transition rates with respect to hierarchical Markov chain with continuous time are uniquely determined by the distributions of their hitting time and sojourn time at bottom states. As a result,the paper can estimate all the statistical characteristics based upon the statistics of the hitting time and sojourn time. The corresponding algorithm and example are provided.