Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Can some one give the technique of this question? $X$ and $Y$ are two independent geometric random variables. $U = \min(X,Y)$ and $V = \max(X,Y)$, $W = V -U$, Find $P(U =i)$ and $P(W =j)$, Also Show that $U$ and $W$ are independent.

GP

share|improve this question

1 Answer 1

up vote 4 down vote accepted

We have $U=\min(X,Y)$; $V=\max(X,Y)$. Also, we have been told, $X$ and $Y$ are geometrically distributed independent random variables! Consider the parameter of the distribution of $X$, equivalently that of $Y$, to be $p$, with $0<p<1$.

Note that the range of $X$ and $Y$ are the set of all whole numbers. This means that, the range of $U$ will be the set of whole numbers as well.

The following quantities are often useful:

$$P(X=i)=p(1-p)^i$$ $$P(X>i)=\sum_{k=i+1}^\infty {p(1-p)^k}=p \cdot\dfrac{(1-p)^{i+1}}{1-(1-p)}=(1-p)^{i+1}$$

Now, use the above to compute the following:

$$P(U=i)=P(\min(X,Y)=i)=P(X>i,Y=i)+P(X=i,Y>i)+P(X=i,Y=i)$$

Leaving all the details of actual computation, [You'll need to use independence and the results I've pointed out previously!], you'll get,

$$P(U=i)=(2p-p^2)\cdot(1-p)^{2i}$$ It is also clear from here that $\min(X,Y)$ follows a geometric distribution with $2p-p^2$ as its parameter. [Note that this parameter is indeed legal as $p^2<2p$.]

Now the second part of the question is interesting: We are interested in the density of $W=V-U$.

$$P(W=j)=P(V-U=j)=\sum_{k=0}^\infty{P(U=k,V=j+k)}$$

Now, if minimum and maximum of two numbers are fixed, essentially both the numbers are known. So, this precisely what is happening here. So, you'll have,

$$\begin{align*}P(W=j)&=\sum_{k=0}^\infty\{P(X=k;Y=j+k)+P(X=j+k;Y=k)\}\\ &=\sum_{k=0}^\infty{2p^2(1-p)^{j+2k}}\\ &=\dfrac{2p^2(1-p)^j}{1-(1-p)^2}\\&=\dfrac{2p^2(1-p)^j}{2p-p^2}\end{align*}$$

Now for the independence, look at the following recipe,

$$\begin{align*}P(U=i;W=j)&=P(U=i;V=i+j)\\&=P(X=i;Y=i+j)+P(X=i+j;Y=i)\end{align*}$$

From here, I assume you'll take through!

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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