Scribd
Upload
3 views

On Certain Discrete Probabilities

The document presents two theorems regarding discrete probabilities on natural numbers. The first theorem establishes the existence of a probability measure for real numbers s > 1, while the second theorem proves that no such measure exists for s in the interval (0, 1]. Both theorems are supported by mathematical proofs involving Riemann's zeta function and properties of prime numbers.

Uploaded by

ronibendavidisrael
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

On Certain Discrete Probabilities

The document presents two theorems regarding discrete probabilities on natural numbers. The first theorem establishes the existence of a probability measure for real numbers s > 1, while the second theorem proves that no such measure exists for s in the interval (0, 1]. Both theorems are supported by mathematical proofs involving Riemann's zeta function and properties of prime numbers.

Uploaded by

ronibendavidisrael
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

On Certain Discrete Probabilities

Theorem 1. For every real number s > 1 there exists a discrete probability P on N, 2 N such that
¡ ¢

P(k N) = k −s

for all k ∈ N.

Proof. Putting P {n} := ζ(s)−1 n −s for all singletons {n} ∈ 2 N , where


¡ ¢

X −s
ζ(s) := n
n≥1

is Riemann’s zeta function, we get a well defined descrete measure P : 2N → R+ which satisfies

P {kn} = ζ(s)−1 k −s
X −s
P(k N) = n = k −s
X ¡ ¢
n≥1 n≥1

as desired, and in particular it holds P(N) = 1, so that P is a probability measure. ä

Theorem 2. Let s ∈ ]0, 1] be a real number. Then there is no probability measure P on N, 2 N such that
¡ ¢

P(k N) = k −s

for all k ∈ N.

Proof. Let P denote the set of all primes and put P n := p ∈ P ¯ n ≤ p as well as P n,m := p ∈ P ¯ n ≤ p ≤m
© ¯ ª © ¯ ª

for all n, m ∈ N. It holds limm→∞ P n,m = P n and limn→∞ P n = ;. We apply a reasoning ad absurdum and
assume that there exists a probability measure P such that P(k N) = k −s . Then we have
 
µ ¶ µ ¶ µ ¶
1−P pN = (−1)card Q P pN = (−1)card Q P p N
[ X \ X Y
p ∈ P n,m Q ⊆ P n,m p ∈Q Q ⊆ P n,m p ∈Q
card Q
p −s 1 −p −s .
X Y Y¡ ¢
= (−1) =
Q ⊆ P n,m p ∈Q p ∈ P n,m

Now recall that, given a sequence (a m )m≥m0 of real numbers a m ∈ ]0, 1[ , we have the equivalence
X Y
a m = ∞ ⇐⇒ (1 − a m ) = 0 .
m ≥m 0 m ≥m 0

But it is well known (and easy to prove) that


p −s = ∞ ,
X
p ∈ Pn

hence
1 −p −s = 0 ,
Y¡ ¢
p ∈ Pn

and it follows ³© µ ¶ µ ¶
ª´
P k ∈ N¯k < n ≤ 1 − P p N = 1 − lim P pN
¯ [ [
m→∞
p ∈ Pn p ∈ P n,m

1 −p −s = 1 −p −s = 0
Y¡ ¢ Y¡ ¢
= lim
m→∞
p ∈ P n,m p ∈ Pn

and finally ³© ª´
1 = P(N) = lim P k ∈ N¯k < n = 0,
¯
n→∞
which is a contradiction. This proves the theorem. ä

You might also like