A1.1 B1.1

Markov Chains | Part II, 2004

(i) Give the definitions of a recurrent and a null recurrent irreducible Markov chain.

Let (Xn)\left(X_{n}\right) be a recurrent Markov chain with state space II and irreducible transition matrix P=(pij)P=\left(p_{i j}\right). Prove that the vectors γk=(γjk,jI),kI\gamma^{k}=\left(\gamma_{j}^{k}, j \in I\right), k \in I, with entries γkk=1\gamma_{k}^{k}=1 and

γik=Ek(# of visits to i before returning to k),ik,\gamma_{i}^{k}=\mathbb{E}_{k}(\# \text { of visits to } i \text { before returning to } k), \quad i \neq k,

are PP-invariant:

γjk=iIγikpij\gamma_{j}^{k}=\sum_{i \in I} \gamma_{i}^{k} p_{i j}

(ii) Let (Wn)\left(W_{n}\right) be the birth and death process on Z+={0,1,2,}\mathbb{Z}_{+}=\{0,1,2, \ldots\} with the following transition probabilities:

pi,i+1=pi,i1=12,i1p01=1\begin{aligned} &p_{i, i+1}=p_{i, i-1}=\frac{1}{2}, i \geq 1 \\ &p_{01}=1 \end{aligned}

By relating (Wn)\left(W_{n}\right) to the symmetric simple random walk (Yn)\left(Y_{n}\right) on Z\mathbb{Z}, or otherwise, prove that (Wn)\left(W_{n}\right) is a recurrent Markov chain. By considering invariant measures, or otherwise, prove that (Wn)\left(W_{n}\right) is null recurrent.

Calculate the vectors γk=(γik,iZ+)\gamma^{k}=\left(\gamma_{i}^{k}, i \in \mathbb{Z}_{+}\right)for the chain (Wn),kZ+\left(W_{n}\right), k \in \mathbb{Z}_{+}.

Finally, let W0=0W_{0}=0 and let NN be the number of visits to 1 before returning to 0 . Show that P0(N=n)=(1/2)n,n1\mathbb{P}_{0}(N=n)=(1 / 2)^{n}, n \geq 1.

[You may use properties of the random walk (Yn)\left(Y_{n}\right) or general facts about Markov chains without proof but should clearly state them.]

Typos? Please submit corrections to this page on GitHub.