0
$\begingroup$

$$\begin{align} p_{ij}(k, k+n) =& \Pr\left\{X_{k+n}=j\mid X_k=i\right\}\\ =& \sum_{r=1}^R \Pr\left\{X_{k+n}=j\mid X_u=r\color{red}, X_k=i\right\}\Pr\left\{X_u=r\mid X_k=i\right\} \end{align}$$

Markov Chains; Using the law of total probability, one can obtain the following equation. I am not entirely sure how it can be obtained. Does the comma between $X_u$ and $X_k$ means intersection between them? $X_u$ is supposed to be a partition of the sample space I think. Any help is appreciated.

$\endgroup$
  • $\begingroup$ $X_u$ is presumably supposed to be a random variable. $\endgroup$ – Nate Eldredge Sep 24 '13 at 23:06
1
$\begingroup$

The comma should mean intersection, as you have guessed.

Remember from the definition of conditional probability,

$$\begin{align} \Pr\left[X_{k+n}=j\mid X_k=i\right] =& \frac{\Pr[X_{k+n}=j, X_k=i]}{\Pr[X_k=i]}\\ =& \frac{\sum_r\Pr[X_{k+n}=j, X_k=i, X_u=r]}{\Pr[X_k=i]}\\ =& \sum_r\frac{\Pr[X_{k+n}=j\mid X_k=i, X_u=r]\Pr[X_k=i, X_u=r]}{\Pr[X_k=i]}\\ =& \sum_r\Pr[X_{k+n}=j\mid X_k=i, X_u=r]\Pr[X_u=r\mid X_k=i] \end{align}$$

where the summation is over all possible value of $X_u$.

$\endgroup$
  • $\begingroup$ Splendid job Peterwhy. The comma does mean intersection as the values of X(u) are summed. Thanks a bunch. $\endgroup$ – qwert87 Sep 24 '13 at 23:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.