2
$\begingroup$

Given two Poisson processes with parameters $\lambda_A$ and $\lambda_B$, I need to find the distribution of events of process A, until the first occurrence of process B.

I realize that the time in question is distributed $\exp(\lambda_B)$ but i'm not sure how to use it.

$\endgroup$
2
$\begingroup$

I assume the two processes are independent.

If you add the two processes together, you get another Poisson process with parameter $\lambda_A+\lambda_B$. If you start with a Poisson process with that parameter, then take the points of that process and label each one of them independently $A$ with probability $\frac{\lambda_A}{\lambda_A+\lambda_B}$ and $B$ with probability $\frac{\lambda_B}{\lambda_A+\lambda_B}$, the points labelled $A$ and those labelled $B$ you get two independent Poisson processes, one with rate $\lambda_A$ and one with rate $\lambda_B$.

This means that if you are told where the points of the combined process occur, each one independently (and independently of where they are) came from process A with probability $\frac{\lambda_A}{\lambda_A+\lambda_B}$. The number of points from process A before the first point from process B is therefore just a geometric random variable with parameter $\frac{\lambda_B}{\lambda_A+\lambda_B}$ (using the definition of geometric where $0$ is a possible value).

| cite | improve this answer | |
$\endgroup$

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.