What is the expected volume of the simplex formed by $n+1$ points independently uniformly distributed on ${\mathbb{S}}^{n-1}$?

beatricalwu
2022-07-17
Answered

What is the expected volume of the simplex formed by $n+1$ points independently uniformly distributed on ${\mathbb{S}}^{n-1}$?

You can still ask an expert for help

uavklarajo

Answered 2022-07-18
Author has **17** answers

Step 1

More generally, for i points independently uniformly distributed in the interior of the n-ball and j points independently uniformly distributed on its boundary (the sphere ${\mathbb{S}}^{n-1}$), with $1\le r:=i+j-1\le n$ so that the points almost surely form an r-simplex, the moments of the volume $\mathrm{\Delta}$ of this simplex are

$E\left[{\mathrm{\Delta}}^{k}\right]=\phantom{\rule{0ex}{0ex}}\frac{1}{r{!}^{k}}{\left(\frac{n}{n+k}\right)}^{i}\frac{\mathrm{\Gamma}(\frac{1}{2}(r+1)(n+k)-j+1)}{\mathrm{\Gamma}(\frac{1}{2}[(r+1)n+rk]-j+1)}{\left(\frac{\mathrm{\Gamma}\left(\frac{1}{2}n\right)}{\mathrm{\Gamma}(\frac{1}{2}[n+k])}\right)}^{r}\prod _{l=1}^{r-1}\frac{\mathrm{\Gamma}(\frac{1}{2}[n-r+k+l])}{\mathrm{\Gamma}(\frac{1}{2}[n-r+l])}\phantom{\rule{thickmathspace}{0ex}}.$

Step 2

In our case, $i=0$, $j=n+1$, $r=n$ and $k=1$, so the desired volume is

${A}_{n}=\frac{1}{n!}\frac{\mathrm{\Gamma}(\frac{1}{2}{n}^{2}+\frac{1}{2})}{\mathrm{\Gamma}\left(\frac{1}{2}{n}^{2}\right)}{\left(\frac{\mathrm{\Gamma}\left(\frac{1}{2}n\right)}{\mathrm{\Gamma}(\frac{1}{2}n+\frac{1}{2})}\right)}^{n}\prod _{l=1}^{n-1}\frac{\mathrm{\Gamma}(\frac{1}{2}l+\frac{1}{2})}{\mathrm{\Gamma}\left(\frac{1}{2}l\right)}\phantom{\rule{thickmathspace}{0ex}}.$

With

$\begin{array}{r}\mathrm{\Xi}(n):=\frac{\mathrm{\Gamma}(n+\frac{1}{2})}{\mathrm{\Gamma}(n)}\end{array}$

this becomes

${A}_{n}=\frac{1}{n!}\mathrm{\Xi}\left(\frac{{n}^{2}}{2}\right)\mathrm{\Xi}{\left(\frac{n}{2}\right)}^{-n}\prod _{l=1}^{n-1}\mathrm{\Xi}\left(\frac{l}{2}\right)\phantom{\rule{thickmathspace}{0ex}}.$

Thus, with

$\begin{array}{ccccccc}n& \frac{1}{2}& 1& \frac{3}{2}& 2& \frac{9}{2}& 8\\ \mathrm{\Xi}(n)& \frac{1}{\sqrt{\pi}}& \frac{\sqrt{\pi}}{2}& \frac{2}{\sqrt{\pi}}& \frac{3\sqrt{\pi}}{4}& \frac{128}{35\sqrt{\pi}}& \frac{6435\sqrt{\pi}}{4096}\end{array}$

we find

${A}_{2}=\frac{1}{2}\frac{\mathrm{\Xi}(2)\mathrm{\Xi}\left(\frac{1}{2}\right)}{\mathrm{\Xi}(1)\mathrm{\Xi}(1)}=\frac{3}{2\pi}$

and

${A}_{3}=\frac{1}{3!}\frac{\mathrm{\Xi}\left(\frac{9}{2}\right)\mathrm{\Xi}\left(\frac{1}{2}\right)\mathrm{\Xi}(1)}{\mathrm{\Xi}\left(\frac{3}{2}\right)\mathrm{\Xi}\left(\frac{3}{2}\right)\mathrm{\Xi}\left(\frac{3}{2}\right)}=\frac{4\pi}{105}\phantom{\rule{thickmathspace}{0ex}},$

in agreement with the MathWorld values, and also

${A}_{4}=\frac{1}{4!}\frac{\mathrm{\Xi}(8)\mathrm{\Xi}\left(\frac{1}{2}\right)\mathrm{\Xi}(1)\mathrm{\Xi}\left(\frac{3}{2}\right)}{\mathrm{\Xi}(2)\mathrm{\Xi}(2)\mathrm{\Xi}(2)\mathrm{\Xi}(2)}=\frac{6435}{31104{\pi}^{2}}\phantom{\rule{thickmathspace}{0ex}}.$

More generally, for i points independently uniformly distributed in the interior of the n-ball and j points independently uniformly distributed on its boundary (the sphere ${\mathbb{S}}^{n-1}$), with $1\le r:=i+j-1\le n$ so that the points almost surely form an r-simplex, the moments of the volume $\mathrm{\Delta}$ of this simplex are

$E\left[{\mathrm{\Delta}}^{k}\right]=\phantom{\rule{0ex}{0ex}}\frac{1}{r{!}^{k}}{\left(\frac{n}{n+k}\right)}^{i}\frac{\mathrm{\Gamma}(\frac{1}{2}(r+1)(n+k)-j+1)}{\mathrm{\Gamma}(\frac{1}{2}[(r+1)n+rk]-j+1)}{\left(\frac{\mathrm{\Gamma}\left(\frac{1}{2}n\right)}{\mathrm{\Gamma}(\frac{1}{2}[n+k])}\right)}^{r}\prod _{l=1}^{r-1}\frac{\mathrm{\Gamma}(\frac{1}{2}[n-r+k+l])}{\mathrm{\Gamma}(\frac{1}{2}[n-r+l])}\phantom{\rule{thickmathspace}{0ex}}.$

Step 2

In our case, $i=0$, $j=n+1$, $r=n$ and $k=1$, so the desired volume is

${A}_{n}=\frac{1}{n!}\frac{\mathrm{\Gamma}(\frac{1}{2}{n}^{2}+\frac{1}{2})}{\mathrm{\Gamma}\left(\frac{1}{2}{n}^{2}\right)}{\left(\frac{\mathrm{\Gamma}\left(\frac{1}{2}n\right)}{\mathrm{\Gamma}(\frac{1}{2}n+\frac{1}{2})}\right)}^{n}\prod _{l=1}^{n-1}\frac{\mathrm{\Gamma}(\frac{1}{2}l+\frac{1}{2})}{\mathrm{\Gamma}\left(\frac{1}{2}l\right)}\phantom{\rule{thickmathspace}{0ex}}.$

With

$\begin{array}{r}\mathrm{\Xi}(n):=\frac{\mathrm{\Gamma}(n+\frac{1}{2})}{\mathrm{\Gamma}(n)}\end{array}$

this becomes

${A}_{n}=\frac{1}{n!}\mathrm{\Xi}\left(\frac{{n}^{2}}{2}\right)\mathrm{\Xi}{\left(\frac{n}{2}\right)}^{-n}\prod _{l=1}^{n-1}\mathrm{\Xi}\left(\frac{l}{2}\right)\phantom{\rule{thickmathspace}{0ex}}.$

Thus, with

$\begin{array}{ccccccc}n& \frac{1}{2}& 1& \frac{3}{2}& 2& \frac{9}{2}& 8\\ \mathrm{\Xi}(n)& \frac{1}{\sqrt{\pi}}& \frac{\sqrt{\pi}}{2}& \frac{2}{\sqrt{\pi}}& \frac{3\sqrt{\pi}}{4}& \frac{128}{35\sqrt{\pi}}& \frac{6435\sqrt{\pi}}{4096}\end{array}$

we find

${A}_{2}=\frac{1}{2}\frac{\mathrm{\Xi}(2)\mathrm{\Xi}\left(\frac{1}{2}\right)}{\mathrm{\Xi}(1)\mathrm{\Xi}(1)}=\frac{3}{2\pi}$

and

${A}_{3}=\frac{1}{3!}\frac{\mathrm{\Xi}\left(\frac{9}{2}\right)\mathrm{\Xi}\left(\frac{1}{2}\right)\mathrm{\Xi}(1)}{\mathrm{\Xi}\left(\frac{3}{2}\right)\mathrm{\Xi}\left(\frac{3}{2}\right)\mathrm{\Xi}\left(\frac{3}{2}\right)}=\frac{4\pi}{105}\phantom{\rule{thickmathspace}{0ex}},$

in agreement with the MathWorld values, and also

${A}_{4}=\frac{1}{4!}\frac{\mathrm{\Xi}(8)\mathrm{\Xi}\left(\frac{1}{2}\right)\mathrm{\Xi}(1)\mathrm{\Xi}\left(\frac{3}{2}\right)}{\mathrm{\Xi}(2)\mathrm{\Xi}(2)\mathrm{\Xi}(2)\mathrm{\Xi}(2)}=\frac{6435}{31104{\pi}^{2}}\phantom{\rule{thickmathspace}{0ex}}.$

asked 2022-08-21

The Probability the number of flips of a fair coin to achieve 3 heads is mod 3

Compute the probability that the number of flips of a fair coin needed in order to achieve a total of exactly 3 Heads will be a multiple of 3.

Compute the probability that the number of flips of a fair coin needed in order to achieve a total of exactly 3 Heads will be a multiple of 3.

asked 2022-08-15

Conditional Probability and Maximum values of random variables including a Geometric Random Variable

Let ${X}_{1},{X}_{2},...$ be i.i.d. random variables with the common CDF F , and suppose they are independent of N, a geometric random variable with parameter p. Let $M=max\{{X}_{1},...,{X}_{N}\}$

(a) Find $Pr\{M\le x\}$ by conditioning on N.

(b) Find $Pr\{M\le x|N=1\}$.

(c) Find $PrM\le x|N>1$.

Could anyone explain why in (c)

$Pr\{M\le x|N>1\}=F(x)Pr\{M\le x\}$

Let ${X}_{1},{X}_{2},...$ be i.i.d. random variables with the common CDF F , and suppose they are independent of N, a geometric random variable with parameter p. Let $M=max\{{X}_{1},...,{X}_{N}\}$

(a) Find $Pr\{M\le x\}$ by conditioning on N.

(b) Find $Pr\{M\le x|N=1\}$.

(c) Find $PrM\le x|N>1$.

Could anyone explain why in (c)

$Pr\{M\le x|N>1\}=F(x)Pr\{M\le x\}$

asked 2022-07-16

Geometric Distribution - transformation $Y=x/(x+1)$

I honestly have no idea where to start with this problem:

Suppose X has the geometric distribution with PMF

$f(x)=\frac{1}{3}{\left(\frac{2}{3}\right)}^{x},x=1,2,.....$

Determine the probability distribution of $Y=x/(x+1).$ Note that both X and Y are discrete random variables. To specify the probability distribution of Y, specify its PMF.

I honestly have no idea where to start with this problem:

Suppose X has the geometric distribution with PMF

$f(x)=\frac{1}{3}{\left(\frac{2}{3}\right)}^{x},x=1,2,.....$

Determine the probability distribution of $Y=x/(x+1).$ Note that both X and Y are discrete random variables. To specify the probability distribution of Y, specify its PMF.

asked 2022-07-16

Probability of geometric Brownian motion

I want to solve the following question: Let X(t) be the price of JetCo stock at time t years from the present. Assume that X(t) is a geometric Brownian motion with zero drift and volatility $\sigma =0.4/yr$. If the current price of JetCo stock is 8.00 USD, what is the probability that the price will be at least 8.40 USD six months from now.

Here is my attempt: Since X(t) is a geometric Brownian motion, log(X(t)) is a regular Brownian motion with zero drift and $\sigma =0.4/yr$. We want to find the probability that $\mathrm{log}(X(1/2))\ge \mathrm{log}(8.40)$ given that $\mathrm{log}(X(0))=\mathrm{log}(8.00)$.

What can I do from here?

I want to solve the following question: Let X(t) be the price of JetCo stock at time t years from the present. Assume that X(t) is a geometric Brownian motion with zero drift and volatility $\sigma =0.4/yr$. If the current price of JetCo stock is 8.00 USD, what is the probability that the price will be at least 8.40 USD six months from now.

Here is my attempt: Since X(t) is a geometric Brownian motion, log(X(t)) is a regular Brownian motion with zero drift and $\sigma =0.4/yr$. We want to find the probability that $\mathrm{log}(X(1/2))\ge \mathrm{log}(8.40)$ given that $\mathrm{log}(X(0))=\mathrm{log}(8.00)$.

What can I do from here?

asked 2022-08-19

A problem on conditional geometric probability

The point (x, y) is chosen randomly in the unit square. What is the conditional probability of ${x}^{2}+{y}^{2}\le 1/4$ given that $xy\le 1/16$.

I started solving this and while calculating I got some very unpleasant numbers but the problem is not marked as difficult in this book where I found it. So I start suspecting that I am having some conceptual mistake.

The two curves intersect at $x=\sqrt{2+\sqrt{3}}/4$ and at $x=\sqrt{2-\sqrt{3}}/4$

Now I have to find a few areas via definite integrals. Is that what this problem is about?

The point (x, y) is chosen randomly in the unit square. What is the conditional probability of ${x}^{2}+{y}^{2}\le 1/4$ given that $xy\le 1/16$.

I started solving this and while calculating I got some very unpleasant numbers but the problem is not marked as difficult in this book where I found it. So I start suspecting that I am having some conceptual mistake.

The two curves intersect at $x=\sqrt{2+\sqrt{3}}/4$ and at $x=\sqrt{2-\sqrt{3}}/4$

Now I have to find a few areas via definite integrals. Is that what this problem is about?

asked 2022-09-19

Tossing a coin. X is the number tosses. What is the probability that X is an odd number?

A "fair" coin is tossed until the first time we get tails. Let X be the number tosses. What is the probability that X is an odd number?

A "fair" coin is tossed until the first time we get tails. Let X be the number tosses. What is the probability that X is an odd number?

asked 2022-08-26

At a given distance from the origin, which convex subsets of an ${\ell}_{p}$-ball have the maximal volume?

For a positive integer n and $p\in [1,\mathrm{\infty}]$, let ${\mathbb{B}}_{n,p}:=\{x\in {\mathbb{R}}^{n}\mid \Vert x{\Vert}_{p}\le 1\}$ be the ${\ell}_{p}$ unit-ball in ${\mathbb{R}}^{n}$. Fix $h\in [0,1]$.

Of all convex subsets A of ${\mathbb{B}}_{n,p}$ with $d(0,A)=h$, which shape maximizes the volume ?

For a positive integer n and $p\in [1,\mathrm{\infty}]$, let ${\mathbb{B}}_{n,p}:=\{x\in {\mathbb{R}}^{n}\mid \Vert x{\Vert}_{p}\le 1\}$ be the ${\ell}_{p}$ unit-ball in ${\mathbb{R}}^{n}$. Fix $h\in [0,1]$.

Of all convex subsets A of ${\mathbb{B}}_{n,p}$ with $d(0,A)=h$, which shape maximizes the volume ?