0
$\begingroup$

I recently started reading Wolfgang Woess' book titled "Random Walks on Infinite Groups". In the section where he introduces Markov chains and random walks on a set $X$, he has defined a probability measure $\mathbb{P}_x$ on $X^{\mathbb{N}_0}$ corresponding to a Markov chain $(Z_n)_{n\geq 0}$ starting at $x\in X$. He has also written how $\mathbb{P}_x$ operates:

$$\mathbb{P}_x[Z_n=x_0, Z_1=x_1,....,Z_n=x_n]=\delta_x(x_0)p(x_0,x_1)\dots p(x_{n-1}, x_n)$$

where $(p(x,y))_{x,y\in X}$ is the transition matrix of the random walk $(Z_n)_{n\geq 0}$.

Question 1: I don't really understand how to apply $\mathbb{P}_x$ on a general measurable set from this given information. For instance, how to calculate $\mathbb{P}_x(Z_m=z)$? More generally how to calculate $\mathbb{P}_x(Y=y)$ where $Y:X^{\mathbb{N}_0}\rightarrow X$ is a random variable ?

Question 2: How does $\mathbb{P}_x$ relate to $(Z_n)_{n\geq 0}$ as $x$ varies?

I am a novice in probability theory so I am extremely sorry in advance if some of my doubts seem absurd. Thanks in advance!

$\endgroup$
2
  • $\begingroup$ Write $\mathbb{P}_x(Z_m=z) = \sum_{x_0,\dots, x_{m-1} \in X} \mathbb{P}_x(Z_0=x_0,Z_1=x_1,\dots,Z_{m-1} = x_{m-1}, Z_m=z)$. $\endgroup$ Commented Feb 21 at 19:09
  • 1
    $\begingroup$ These questions are quite general and quite basic - to address them, how about starting by reading an introductory text on discrete-time Markov chains? For example, Chapter 1 of James Norris's book here: statslab.cam.ac.uk/~james/Markov (but many other texts with different styles are available). Then you'll be well placed to ask more focused questions about anything you don't understand. (For that purpose math.stackexchange.com will probably be more suitable than MathOverflow.) $\endgroup$ Commented Feb 21 at 22:05

0

You must log in to answer this question.