Revision as of 00:13, 6 December 2020 by Nlfang (Talk | contribs)


Communication and Reducibility

Before moving on and discovering more behaviors of Markov chains, we need to classify the properties of states first. In the previous section, we use $ P_{ij} $ to represent the transition probability of state changing from $ i $ to $ j $; now, we use $ P_{ij}^{n} $ to denote the transition probability of state changing from $ i $ to $ j $ after $ n $ step. If $ P_{ij}^{n} \geq 0 $ for all $ n \geq 1 $, we say state $ j $ is accessible from state $ i $, which also can be written as $ i \rightarrow j $. If $ i \rightarrow j $ and at the same time $ j \rightarrow i $, then we say that state $ i $ and state $ j $ communicate, which can be denoted as $ i \leftrightarrow j $. On the other hand, if $ P_{ij}^{n} = 0 $ or $ P_{ji}^{n} = 0 $, or if both of them come into existence, we claim that state $ i $ and state $ j $ do not communicate. The diagrams below illustrate these two cases.

Markovcommunicate.png
Irreduciblemarkov.png
Reduciblemarkov1.png
Reduciblemarkov2.png

Back to Markov Chains

Alumni Liaison

Prof. Math. Ohio State and Associate Dean
Outstanding Alumnus Purdue Math 2008

Jeff McNeal