Hey could someone explain why this is true? I am trying to understand how to solve such a problem but I dont understand the solution.
Given a Markov chain [itex]\left\{X_{n}: n\in\ \mathbb{N}\right\}[/itex] with four states 1,2,3,4 and transition matrix
[itex]P =
\begin{pmatrix}
0 & \frac{1}{2}& \frac{1}{2} & 0 \\
0 & 0 & \frac{1}{2} & \frac{1}{2}\\
\frac{1}{2} & 0 & 0 & \frac{1}{2} \\
\frac{1}{2} & \frac{1}{2} & 0 & 0
\end{pmatrix}[/itex]
We leave from state 3. What is the chance that state 1 will be reached before state 4?
Solution:
Lets call this chance x. Lets call y the same chance but instead leaving from 2. Lets call:
[itex]f =
\begin{pmatrix}
1 \\
x \\
y \\
0
\end{pmatrix}[/itex]
Then f satisfies P'f = f with P' the transition matrix derived from P by making state 1 and 4 absorbing states i.e.,
[itex]
\begin{pmatrix}
1 & 0& 0 & 0 \\
0 & 0 & \frac{1}{2} & \frac{1}{2}\\
\frac{1}{2} & 0 & 0 & \frac{1}{2} \\
0 & 0 & 1 & 1\end{pmatrix}*\begin{pmatrix}
1 \\
x \\
y \\
0
\end{pmatrix}= \begin{pmatrix}
1 \\
x \\
y \\
0
\end{pmatrix}[/itex]
Solving gives y = 1/2, x = 1/4.
Given a Markov chain [itex]\left\{X_{n}: n\in\ \mathbb{N}\right\}[/itex] with four states 1,2,3,4 and transition matrix
[itex]P =
\begin{pmatrix}
0 & \frac{1}{2}& \frac{1}{2} & 0 \\
0 & 0 & \frac{1}{2} & \frac{1}{2}\\
\frac{1}{2} & 0 & 0 & \frac{1}{2} \\
\frac{1}{2} & \frac{1}{2} & 0 & 0
\end{pmatrix}[/itex]
We leave from state 3. What is the chance that state 1 will be reached before state 4?
Solution:
Lets call this chance x. Lets call y the same chance but instead leaving from 2. Lets call:
[itex]f =
\begin{pmatrix}
1 \\
x \\
y \\
0
\end{pmatrix}[/itex]
Then f satisfies P'f = f with P' the transition matrix derived from P by making state 1 and 4 absorbing states i.e.,
[itex]
\begin{pmatrix}
1 & 0& 0 & 0 \\
0 & 0 & \frac{1}{2} & \frac{1}{2}\\
\frac{1}{2} & 0 & 0 & \frac{1}{2} \\
0 & 0 & 1 & 1\end{pmatrix}*\begin{pmatrix}
1 \\
x \\
y \\
0
\end{pmatrix}= \begin{pmatrix}
1 \\
x \\
y \\
0
\end{pmatrix}[/itex]
Solving gives y = 1/2, x = 1/4.
via Physics Forums RSS Feed http://www.physicsforums.com/showthread.php?t=703415&goto=newpost
0 commentaires:
Enregistrer un commentaire