MHB A question related to cardinality and probability

Click For Summary
The discussion centers on a question about the relationship between the expected values of intersections of two sets, A and B, with a subset of incidents I. Given that A and B have equal sizes and that the intersection of A with a subset I1 has a greater or equal cardinality than that of B, the main inquiry is whether the expected value of the intersection of A with another subset I2 is also greater than or equal to that of B. Participants clarify the notation and assumptions, emphasizing the need for rigorous proof or counterexamples to support or refute the proposed statement. The conversation highlights the complexities of probability and cardinality in set theory.
baiyang11
Messages
8
Reaction score
0
Dear all,

I have a question attached related to both probability and cardinality. Let me know if my formulation of the problem is non-rigorous or confusing. Any proof or suggestions are appreciated.Thank you all.
The question follows.Consider a set \(I\) consists of \(N\) incidents.
\[I=\{i_{1},i_{2},...,i_{k},...i_{N}\}\]
Each incident has a probability to happen, i.e. incident \(i_{k}\) happens with the probability \(r_{k}\). Without loss of generality, we assume \(r_{1}\geq r_{2}\geq ... \geq r_{k}\geq ... \geq r_{N}\)
Given a constant \(n<N\), we can have set \(I_{1}=\{i_{1},i_{2},...,i_{n}\}\). Apparently, \(|I_{1}|=n\) and \(I_{1}\subset I\).
Define a mapping \(I\to S\) with \(S=\{s_{1},s_{2},...,s_{k},...s_{N}\}\) subject to
\[
s_{k} = \left\{ \begin{array}{ccc}
1 &\mbox{ (Pr=$r_{k}$)} \\
0 &\mbox{ (Pr=$1-r_{k}$)} \\
\end{array} \right.
\]
Pick out the incidents with correspond \(s\) being 1 to form the set \(I_{2}\) , i.e.
\[I_{2}=\{i_{m_{1}},i_{m_{2}},...,i_{m_{M}}\} \quad \mbox{and} \quad s_{m_{k}}=1 \quad k=1,2,...,M \]
Apparently, \(|I_{2}|=M\) and \(I_{2}\subset I\). Note that there could be \(I_{2}\ne I_{1}\) and \(|I_{2}| \ne |I_{1}|\).

The question is,
If we have two set \(A\) and \(B\) with \(|A|=|B|=n\) and assume
\[ |A \cap I_{1}| \geq |B \cap I_{1}| \]
Is the following statement true?
\[ E(|A \cap I_{2}|) \geq E(|B \cap I_{2}|) \]
where \(E\) means expected value.
If this is true, how to prove it? If not, how to prove it’s not true?
 
Physics news on Phys.org
I already looked a couple of times at this question and it still confuses me.
First of all the statement itself. I think it has to be stated as
$$\mathbb{E}(A \cap I_2) \geq \mathbb{E}(B \cap I_2)?$$

Second, is there no information about $A$ en $B$ at all? Perhaps, something like $A \subset I$ and $B \subset I$?
 
Siron said:
I already looked a couple of times at this question and it still confuses me.
First of all the statement itself. I think it has to be stated as
$$\mathbb{E}(A \cap I_2) \geq \mathbb{E}(B \cap I_2)?$$

Second, is there no information about $A$ en $B$ at all? Perhaps, something like $A \subset I$ and $B \subset I$?

Thank you for reply!
(1) Yes. Here the 'E' character means expected value. I just didn't know how to make it like your character when I wrote this question. Now I know, but I can't edit it.
(2) Yes. $A \subset I$ and $B \subset I$.
Since I can't edit the original post, let me repost the question here.

Consider a set $I$ consists of $N$ incidents.
$I=\{i_{1},i_{2},...,i_{k},...i_{N}\}$
Each incident has a probability to happen, i.e. incident $i_{k}$ happens with the probability $r_{k}$. Without loss of generality, we assume $r_{1}\geq r_{2}\geq ... \geq r_{k}\geq ... \geq r_{N}$
Given a constant $n<N$, we can have set $I_{1}=\{i_{1},i_{2},...,i_{n}\}$. Apparently, $|I_{1}|=n$ and $I_{1}\subset I$.
Define a mapping $I\to S$ with $S=\{s_{1},s_{2},...,s_{k},...s_{N}\}$
subject to
$s_{k} = \left\{ \begin{array}{ccc}
1 &\mbox{ (Pr=$r_{k}$)} \\
0 &\mbox{ (Pr=$1-r_{k}$)} \\
\end{array} \right.$
Pick out the incidents with correspond $s$ being 1 to form the set $I_{2}$ , i.e.

$I_{2}=\{i_{m_{1}},i_{m_{2}},...,i_{m_{M}}\} \quad \mbox{and} \quad s_{m_{k}}=1 \quad k=1,2,...,M$

Apparently, $|I_{2}|=M$ and $I_{2}\subset I$. Note that there could be $I_{2}\ne I_{1}$ and $|I_{2}| \ne |I_{1}|$.

The question is,
If we have two set $A$ and $B$ satisfying the following three assumptions:

(1)$A\subset I$ and $B\subset I$

(2)$|A|=|B|=n$

(3)$|A \cap I_{1}| \geq |B \cap I_{1}|$

Is the following statement true?

$\mathbb{E}(A \cap I_2) \geq \mathbb{E}(B \cap I_2)$

where $\mathbb{E}$ means expected value.
If this is true, how to prove it? If not, how to prove it’s not true?
 
Last edited:
Greetings, I am studying probability theory [non-measure theory] from a textbook. I stumbled to the topic stating that Cauchy Distribution has no moments. It was not proved, and I tried working it via direct calculation of the improper integral of E[X^n] for the case n=1. Anyhow, I wanted to generalize this without success. I stumbled upon this thread here: https://www.physicsforums.com/threads/how-to-prove-the-cauchy-distribution-has-no-moments.992416/ I really enjoyed the proof...

Similar threads

  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 11 ·
Replies
11
Views
1K
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
0
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 21 ·
Replies
21
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
Replies
2
Views
2K