Computing $prob(x\ge 4|x\ge 1)$ using geometric distribution

50 Views Asked by At

I am wondering if I did this right. Compute the following

Computing

$prob(x\ge 4|x\ge 1)$

using the geometric distribution

The geometric distribution is $f(x)=p(1-p)^x$

So then would I do the following

$prob(x\ge 4|x\ge 1)=prob(x \ge 4-1)$

by the memory less property so then find $prob(x \ge 3)$

and I get

$\int_3^{\infty}p(1-p)^x$

$=\frac{p(1-p)^x}{\log(1-p)}$ from 3 to infinity and then I get

$=\frac{(p-1)^3}{\log(1-p)}$

1

There are 1 best solutions below

0
On

In general it is a good idea to use the converse probability to calculate $P(X\geq x)$ if $x$ is not large-as suggested by Henry. It works perfectly here. But in case of the geometric distribution the expression for $P(X\geq x)$ is quite simple.

$$P(X\geq x)=\sum_{k=x}^{\infty} p\cdot q^{k}=p\cdot \sum_{k=x}^{\infty}q^{k}$$

$$=p\cdot \left[ \sum_{k=0}^{\infty}q^{k}-\sum_{k=0}^{x-1}q^{k} \right]= p\cdot \left[ \frac1{1-q}-\frac{1-q^x}{1-q} \right]$$

Since $p=1-q$ the denominator disappears by cancelling. Therefore

$$P(X\geq x)=q^x=(1-p)^x$$

For $p=0.3$ and $x=3$ we can check both ways:

$\texttt{a) closed forumula}$

$P(X\geq 3)=(1-0.3)^3=0.7^3$

$\texttt{b) converse probability}$

b) $P(X\geq 3)=1-P(X=2)-P(X=1)-P(X=0)=1-0.3\cdot 0.7^2-0.3\cdot 0.7-0.3$

$$=0.343=0.7^3$$