|
|
| Author |
Message |
yangyu05@gmail.com science forum beginner
Joined: 28 Oct 2005
Posts: 7
|
Posted: Thu Feb 23, 2006 5:40 pm Post subject:
A stochastic geometry problem
|
|
|
Consider a circular field with k independently and uniformaly
distributed points. Suppose we randomly drop one more point, A, in the
field. Let the distance from A to the k points, d(A), be the shortest
distance between A and all k points. Can we get the expected value of
d(A)?
Thanks! |
|
| Back to top |
|
 |
One Usenet Poster science forum beginner
Joined: 03 May 2005
Posts: 9
|
Posted: Thu Feb 23, 2006 9:09 pm Post subject:
Re: A stochastic geometry problem
|
|
|
yang wrote:
| Quote: | Consider a circular field with k independently and uniformaly
distributed points. Suppose we randomly drop one more point, A, in the
field. Let the distance from A to the k points, d(A), be the shortest
distance between A and all k points. Can we get the expected value of
d(A)?
Thanks!
|
0.5 * sqrt( area / (k+1) ) |
|
| Back to top |
|
 |
john2 science forum beginner
Joined: 01 Feb 2006
Posts: 15
|
Posted: Fri Feb 24, 2006 1:09 pm Post subject:
Re: A stochastic geometry problem
|
|
|
yang wrote:
| Quote: | Consider a circular field with k independently and uniformaly
distributed points. Suppose we randomly drop one more point, A, in the
field. Let the distance from A to the k points, d(A), be the shortest
distance between A and all k points. Can we get the expected value of
d(A)?
Thanks!
|
If A is not too close to the edge of the field:
Consider a circle, centre A, radius r.
The mean number of points in the circle is
mu = pi.r^2.sigma
where sigma is the point density = field area/k.
Probability of no points inside r is given by the Poisson result
P_0(r) = exp(- mu)
This is the same probability as the nearest point being at distance r
(from rank statistics)
Next differentiate P_0(r) wrt r to find the pdf of the nearest point.
Find the expectation of this pdf.
When A is close to the edge of the field it will get a *lot* more
complicated
john2 |
|
| Back to top |
|
 |
Mike science forum addict
Joined: 17 Sep 2005
Posts: 74
|
Posted: Sat Feb 25, 2006 11:41 pm Post subject:
Re: A stochastic geometry problem
|
|
|
"One Usenet Poster" <me@my.computer.org> wrote in message
news:11vs93h1vokrafd@corp.supernews.com...
| Quote: | yang wrote:
Consider a circular field with k independently and uniformaly
distributed points. Suppose we randomly drop one more point, A, in the
field. Let the distance from A to the k points, d(A), be the shortest
distance between A and all k points. Can we get the expected value of
d(A)?
Thanks!
0.5 * sqrt( area / (k+1) )
Could you please explain how you got this result. Thanks. |
|
|
| Back to top |
|
 |
One Usenet Poster science forum beginner
Joined: 03 May 2005
Posts: 9
|
Posted: Sun Feb 26, 2006 9:18 pm Post subject:
Re: A stochastic geometry problem
|
|
|
Mike wrote:
| Quote: | "One Usenet Poster" <me@my.computer.org> wrote in message
news:11vs93h1vokrafd@corp.supernews.com...
yang wrote:
Consider a circular field with k independently and uniformaly
distributed points. Suppose we randomly drop one more point, A, in the
field. Let the distance from A to the k points, d(A), be the shortest
distance between A and all k points. Can we get the expected value of
d(A)?
Thanks!
0.5 * sqrt( area / (k+1) )
Could you please explain how you got this result. Thanks.
|
I didn't derive it; I just found this
www.css.cornell.edu/courses/620/lecture11.ppt via Google.
The general expression for the k_th order is:
d(k) = k * (2 * k)! / ( (2^k k!)^2 sqrt( N / area ) )
So for k = 1, we get the above result.
Mike |
|
| Back to top |
|
 |
Google
|
|
| Back to top |
|
 |
|
|
The time now is Thu Jan 08, 2009 2:04 pm | All times are GMT
|
|
Loans | Mortgage Calculator | Mobile Phones | Mortgages | Guitar Books
|
|
Copyright © 2004-2005 DeniX Solutions SRL
|
|
Other DeniX Solutions sites:
Electronics forum |
Medicine forum |
Unix/Linux blog |
Unix/Linux documentation |
Unix/Linux forums
|
Powered by phpBB © 2001, 2005 phpBB Group
|
|