Steady-State Approximation for a Vector Valued Markov Chain
针对马尔可夫链 (X(t), S(t)) 提出一种迭代算法，在已知 S(t) 稳态边缘分布的情况下，近似求解 X(t) 的稳态边缘分布，适用于 S(t) 仅为满足马尔可夫假设而引入的辅助变量。
The problem addressed is that of a condensed steady-state solution for the Markov Chain (X(t), S(t)). The steady state marginal distribution of S(t) is known; we desire only the steady state marginal distribution of X(t). Such a case frequently arises when the supplementary random variable S(t) is required in the state description solely to satisfy the Markovian assumption. An iterative algorithm is presented which makes use of an approximation to the conditional distribution for S(t) given X(t).