TWO THEOREMS ON CONVERGENCE PARAMETER OF AN IRREDUCIBLE MARKOV CHAIN

被引:0
|
作者
Shur, M. G. [1 ]
机构
[1] Moscow State Inst Elect & Math, Moscow 109028, Russia
关键词
irreducible Markov chain; symmetric Markov chain; convergence parameter;
D O I
10.1137/S0040585X9798645X
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
A homogeneous irreducible Markov chain X with a measurable state space (E, B) and a transient operator P acting in a probability of bounded from below measurable functions is considered. The sigma-algebra beta is assumed to be countably generated. It is proved that if the chain is aperiodic and function f and measure. are small, then [nu(P(n)f)](1/n) -> R as n -> infinity, where R is the convergence parameter. For periodic Markov chains this statement can be modified in the following way. If a chain X is symmetric with respect to some sigma-finite measure pi, then R = vertical bar vertical bar(P) over tilde vertical bar vertical bar(-1), where (P) over tilde is a bounded self-adjoint operator generated by P and acting in the space L-2(pi). Results of this paper extend the results of M. G. Shur
引用
收藏
页码:159 / U177
页数:6
相关论文
共 50 条