Old Dominion University on - NASA2011 Slide Competition For The Health Risks of Extraterrestrial...

Post on 14-Oct-2020

1 views 0 download

Transcript of Old Dominion University on - NASA2011 Slide Competition For The Health Risks of Extraterrestrial...

The Annual NASA Space Radiation Summer School 2011 Slide Competition

For The Health Risks of Extraterrestrial Environments (THREE)

Honorable MentionCandice Rockell GerstnerOld Dominion University

Submission onThe Use of Poisson Distribution for Cell Hits

fromLecture on Radiation Interactions with Matter

Tom Borak (2011)

The Use of Poisson Distribution for Cell Hits

Dr. Candice Rockell Gerstner

Old Dominion University, Norfolk, VA 23529crockell@odu.edu

1/7

Chosen Slide from Dr. Tom Borak

kTrack structure

)( nn

!ne)n(]p[n

n-nH

H

HH

!nH

2/7

Derivation of Poisson Distribution

In this slide, Dr. Borak was showing the equation of how to calculate theprobability of hitting nH cells. This probability is given by a PoissonDistribution, which is shown in his slide. I would now like to discuss thederivation of Poisson’s Distribution.

First, let’s start out with a few facts that we know:1 - lim

n→∞(1 + c

n )n = expc

2 - From combinatorics: C (n, k) = n!k!(n−k)! .

3 - limn→∞

C(n,k)nk = lim n→∞ n!

k!(n−k)!nk . By reducing the fraction and using

limit laws, it can be found that this limit is equal to 1k! .

3/7

Derivation of Poisson Distribution

4 - Binomial Distribution is used to approximate the distribution of arandom variable X with k successes (hits) out of n trials (cells), whereP(X = k) = C (n, k)pkqn−k . In this definition, p is the probability ofsuccess and q is the probability of failure. However, this computation canbe very tedious and it is also known that there is a small probability ofmultiple hits that is not taken into account with this formula.

4/7

Derivation of Poisson Distribution

Suppose that the average number of particles arriving in a interval oflength/second is γ so that the distribution of a random variable X withthe expectation is E (X ) = γ.

If small intervals of time are taken, we can then consider that theprobability of success in a binomial distribution with n trials isP(success) = γ

n , where n is the number of trials. Therefore,

P(X = k) = limn→∞

C (n, k)(γ

n)k(1− γ

n)n−k

= γk limn→∞

C (n, k)(1

n)k(1− γ

n)n−k . (1)

5/7

Derivation of Poisson Distribution

Using the limits that we stated in numbers 1-4,

P(X = k) = γk(1

k!) lim

n→∞(1 +

−γn

)n

= γk(1

k!) exp−γ

=(ave. number of hits)hits · exp−ave. number of hits

hits!(2)

Fact: The Poisson Distribution was developed by a French mathematicianof the nineteenth century. Some applications include radioactive counts,the number of cells that divide in a particular unit of time, number ofdefects in a length of wire, etc.

6/7