How is the intersection of two independent sets such as $A$ and $B$ a multiplication process? That is, why is $P(A \cap B)=P(A)P(B)$?

  • $\begingroup$ Problem is quite unclear. Total answers of 2-D matrix product? Independent events? What is it all about? And did you want to understand these in K.G? $\endgroup$ – Ganesh Oct 20 '12 at 22:44

Why does $P(A)\cap P(B)=P(A)\times P(B) $ for independent events $A$ and $B$?

Your original question noted that multiplication usually computes the number of combinations, and you can model probabilities in a similar manner. Suppose you have two events, $A$ and $B$, such that $A$ occurs 40% of the time, and $B$ occurs 60% of the time. We can make a matrix of possible outcomes by multiplying these probabilities by some number of occurrences. Then we look to see which combinations are the event $(A,B)$.

Probability Matrix

We see that 6 of the 25 occurrences are $(A,B)$; 24% of all occurrences. We could have computed this by multiplying the number of $A$ by the number of $B$ and dividing by the total number of occurrences: $$\frac{\#A\times \#B}{(\#A+\#\neg A)\times(\#B+\#\neg B)}=\frac{\#A}{(\#A+\#\neg A)}\frac{\#B}{(\#B+\#\neg B)}=P(A)P(B)$$


There is a universe $\Omega$ of possible outcomes (or histories) $\omega$. Any reasonable subset $A\subset\Omega$ is called an event and is assigned a certain probability $P(A)\in[0,1]$. Two events $A$ and $B$ are called independent if $$P(A\cap B)=P(A)\,P(B)\ .\qquad(*)$$ So what you call a "multiplication process" is actually the definition of independence.

Of course this definition is derived from our intuition about "independence": Knowing that $B$ has indeed happened should not alter the prospects of the occurrence of $A$. This means that the fraction of outcomes $\omega\in A$ among all outcomes $\omega\in B$ should be the same as the fraction of outcomes $\omega\in A$ among all outcomes $\omega\in\Omega$. Written as a formula this is $${P(A\cap B)\over P(B)}={P(A)\over P(\Omega)}=P(A)\ ,$$ which coincides with $(*)$ if $P(B)\ne0$.


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.