2
$\begingroup$

In a physics book, I've come across the quantity

$$ \frac{\def\P{\mathbb{P}}\P(A\cap B)}{\P(A)\P(B)}\,, $$

where $A$ and $B$ are events.

The author calls this quantity the correlation of $A$ and $B$, but the expression above does not agree with the definition of "correlation" that I'm familiar with. (For one thing, when $A$ and $B$ are independent, the quantity above equals $1$, yet the correlation I'm familiar with is $0 \neq 1$ when the events are independent.)

I must conclude that the author's word choice reflects the usage within physics.

Does this quantity have a standard name in mathematics?

$\endgroup$
  • $\begingroup$ not that I have known. And I am a graduate in probability... $\endgroup$ – mookid Apr 10 '14 at 23:42
2
$\begingroup$

I think the book is simply saying that it is a measure of how correlated the two variables are. If they are totally independent (un-correlated) then $P(A\cap B)=P(A)P(B)$ and the quantity is $1$. If they are mutually exclusive the quantity is $0$. In general, $$\frac{P(A\cap B)}{P(A)P(B)}=\frac{P(A|B)}{P(A)}=\frac{P(B|A)}{P(B)}$$ So it measures what proportion of $A$ occurs "in" $B$ or vice versa.

But to answer the question, I think it's fair to say that 1) it almost certainly has some name and 2) that name is almost certainly not standard. So no - it has no name.

$\endgroup$
1
$\begingroup$

There are a number of dependence measures which are defined in similar way. For example, given two $\sigma$-fields $F$ and $G$:

$\alpha(F,G)=sup_{(A\in F,B\in G)}|P(A\cap B)-P(A)P(B)|$

$\phi(F,G)=sup_{(A\in F,B\in G)}|P(A|B)-P(A)|$

Your dependence measure defined using similar logic:

$\delta(A,B)=\frac{P(A\cap B)}{P(A)P(B)}=\frac{P(A|B)}{P(A)}$

  • If events $A$ and $B$ are independent $\delta(A,B)=1$

  • If events $A$ and $B$ are dependent than $P(A|B)>(<)P(A)$ and $\delta(A,B)>(<)1$

$\endgroup$
1
$\begingroup$

I think "pointwise mutual information" is what comes close. PMI is defined as

$\log{\frac{p(x,y)}{p(x)p(y)}}$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.