0
$\begingroup$

Consider the random variables $Y_1,Y_2,...,Y_n$ which are independent random variables such that $Y_i \sim U[0,\theta]$ if $i$ is even and $Y_i \sim U[1,2\theta]$ is $i$ is odd. I've been trying to figure out the probability distribution function for $$Y_{(n)} = \min\{Y_1,Y_2,...,Y_n\}$$ Like standard order statistics, I start out by calculating the CDF as $$f_{Y_{(n)}} = P(Y_{(n)} \leq y) = 1 - P(Y_{(n)} > y) = 1-P(Y_1 > y \wedge Y_2 > y \wedge ... \wedge Y_n > y)$$, which due to independence thankfully allows for a product to form: $$1-\prod_{i=1}^n P(Y_i > y).$$ However, I am having difficulty calculating this product given the different distributions among the random variables that I need to consider. Any suggestions on this computation?

$\endgroup$
1
$\begingroup$

Not sure if you consider the following floor notation as "computation" - you just merely need to count the number of odd and even numbers in the first $n$ natural numbers. The last result can be expressed as $$ F_{Y_{(1)}}(y) = \begin{cases} 0 & y \in (-\infty, 0) \\ \displaystyle 1 - \left(1 - \frac {y} {\theta}\right)^{\lfloor \frac {n} {2} \rfloor} & y \in (0, 1) \\ \displaystyle 1 - \left(1 - \frac {y} {\theta}\right)^{\lfloor \frac {n} {2} \rfloor} \left(1 - \frac {y-1} {2\theta - 1}\right)^{\lfloor \frac {n+1} {2} \rfloor} & y \in (1, \theta) \\ 1 & y \in (\theta, +\infty) \end{cases} $$

$\endgroup$

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.