多元马氏模型Phase-type分布的反问题

The Inverse Problem to Phase-type Distributions of a Multivariate Markov Model

本文在多元马尔可夫模型下研究Phase-Type分布的反问题.在确定模型中各随机序列的瞬时状态集转移到吸收态集的首达时间的条件下,根据多元马氏模型的首达时间的条件分布向量序列,针对PhaseType分布的反问题,求出多元马氏模型瞬时状态集的转移概率矩阵.

In this paper we consider the inverse problem to phase- type distributions of a multivariate Markov model. Provided that the hitting time from the transient state set of the stochastic sequences to the absorbing state set is determined,and then using the conditional distribution of vector sequence in the multivariate Markov model,the inverse problem to phase- type distributions is solved,and the transition probability matrix of transient state set for the multivariate Markov model is obtained.