The middle number

Given 5 5 numbers, each randomly chosen between zero and one, what is the probability that the number in the "middle" (the third one when sorted) is between 0.4 0.4 and 0.6 0.6 ?

If this probability is a b \dfrac{a}{b} where a a and b b are coprime positive integers, what is a + b a+b ?


The answer is 4266.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Mark Hennings
Jun 13, 2019

If 0 < x < 1 0 < x < 1 and δ x \delta x is an infinitesimal, the probability that the middle number X X lies between x x and x + δ x x + \delta x is P [ x x x + δ x ] = 5 × ( 4 2 ) × x 2 × ( 1 x δ x ) 2 × δ x 30 x 2 ( 1 x ) 2 δ x P[x \le x \le x + \delta x] \; = \; 5 \times \binom{4}{2} \times x^2 \times (1-x - \delta x)^2 \times \delta x\; \approx \; 30x^2(1-x)^2\delta x (two of the numbers must be less than x x , two must be greater than x + δ x x + \delta x , one must be between x x and x + δ x x + \delta x , and there are 5 × ( 4 2 ) = 30 5 \times \binom{4}{2} = 30 ways in which this can happen). Thus the probability density function of X X is f X ( x ) = { 30 x 2 ( 1 x ) 2 0 < x < 1 0 o . w . f_X(x) \; = \; \left\{ \begin{array}{lll} 30x^2(1-x)^2 & \hspace{1cm} & 0 < x < 1 \\ 0 & & \mathrm{o.w.}\end{array}\right. and hence the desired probability is P [ 2 5 < X < 3 5 ] = 30 2 5 3 5 x 2 ( 1 x ) 2 d x = 1141 3125 P[\tfrac25 < X< \tfrac35] \; = \; 30\int_{\frac25}^{\frac35} x^2(1-x)^2\,dx \; = \; \tfrac{1141}{3125} making the answer 1141 + 3125 = 4266 1141 + 3125 = \boxed{4266} .

Nice write up, Mark!

Geoff Pilling - 1 year, 12 months ago

In fact, the k th k^\text{th} order statistic of a uniform distribution follows a beta distribution.

U ( k ) Beta ( k , n + 1 k ) U_{(k)} \sim \text{Beta} (k, n + 1 - k) . In this case, n = 5 , k = 3 n = 5, k = 3 .

Pi Han Goh - 1 year, 11 months ago

Log in to reply

Interesting observation, Pi!

Geoff Pilling - 1 year, 11 months ago

You have the good result but your writing seems false. Indeed you look at one specifique way where there is two numbers < 0.4 & two numbers > 0.6, but there are many more. Bellow, there is a list of all posibility in a tree, "<" means less than 0.4, ">" means more than 0.6, "=" means between 0.4 and 0.6 : ===== appears 1 time, probability 0.2^5 <==== appears 5 times, probability 0.2^4 * 0.4 ====> appears 5 times, probability 0.2^4 * 0.4 <<=== appears 10 times, probability 0.2^3 * 0.4^2 ===>> appears 10 times, probability 0.2^3 * 0.4^2 <==>> appears 30 times, probability 0.2^2 * 0.4^3 <<==> appears 30 times, probability 0.2^2 * 0.4^3 <===> appears 20 times, probability 0.2^3 * 0.4^2 <<=>> appears 30 times, probability 0.2 * 0.4^4 there is your one 30 * 0.4^2 * (1 - 0,4 - 0,2)^2 (Note that you can use a binary tree with the case where it's between and the case where it's not). Then multiply for each posibility, the time it appears by the probability. Finaly sum each probability you have obtained just before and you obtain 0,36512 which is egal to 1141/3125

Thomas El Haj Hussein - 1 year, 10 months ago

Log in to reply

We are talking about continuous random variables, so the probability that any one of these takes a specific value is 0 0 . Thus the probability of being less than 1 1 , for example, is the same as the probability of being less than or equal to 1 1 .

Mark Hennings - 1 year, 10 months ago

Did everything you did except I forgot the 30 in the calculation. :(

Razzi Masroor - 1 year, 5 months ago
Yuriy Kazakov
Oct 21, 2020

Calculate the probability P ( m , n , k ) P(m,n,k) -here m m - numbers are between 0 0 and 0.4 0.4 , n n - numbers are between 0.4 0.4 and 0.6 0.6 , k k - numbers are between 0.4 0.4 and 0.6 0.6 , m + n + k = 5 m+n+k=5 .

P ( 2 , 1 , 2 ) = C 5 2 C 3 1 2 4 5 4 1 5 = 480 5 5 P(2,1,2)=C^2_{5}C^1_{3} {\frac{2^4}{5^4}} \frac{1}{5}=\frac{480}{5^5}

P ( 1 , 3 , 1 ) = C 5 1 C 4 1 2 2 5 2 1 5 3 = 80 5 5 P(1,3,1)=C^1_{5}C^1_{4} {\frac{2^2}{5^2}} \frac{1}{5^3}=\frac{80}{5^5}

P ( 2 , 2 , 1 ) = P ( 1 , 2 , 2 ) = C 5 1 C 4 2 2 3 5 3 1 5 2 = 240 5 5 P(2,2,1)=P(1,2,2)=C^1_{5}C^2_{4} {\frac{2^3}{5^3}} \frac{1}{5^2}=\frac{240}{5^5}

P ( 0 , 3 , 2 ) = P ( 2 , 3 , 0 ) = C 5 1 2 2 5 2 1 5 3 = 40 5 5 P(0,3,2)=P(2,3,0)=C^1_{5} {\frac{2^2}{5^2}} \frac{1}{5^3}=\frac{40}{5^5}

P ( 0 , 4 , 1 ) = P ( 1 , 4 , 0 ) = C 5 1 2 5 1 5 4 = 10 5 5 P(0,4,1)=P(1,4,0)=C^1_{5} {\frac{2}{5}} \frac{1}{5^4}=\frac{10}{5^5}

P ( 0 , 5 , 0 ) = 1 5 5 P(0,5,0)= \frac{1}{5^5}

P ( m , n , k ) = 480 5 5 + 80 5 5 + 2 240 5 5 + 2 40 5 5 + 2 10 5 5 + 1 5 5 = 1141 3125 \sum {P(m,n,k)}= \frac{480}{5^5}+\frac{80}{5^5}+2 \cdot\frac{240}{5^5}+2\cdot\frac{40}{5^5}+2\cdot\frac{10}{5^5} +\frac{1}{5^5}=\frac{1141}{3125}

k k - numbers are between 0.6 0.6 and 1.0 1.0

Yuriy Kazakov - 7 months, 3 weeks ago

the combinatorial coefficients are counting how many ways to assign indices 1,2,3,4,5 to the different partitions (i.e. 2,1,2 is a partition into a group of 2, one group of 1, and a 3rd group of 2)?

Nathan Zhao - 5 months, 3 weeks ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...