3
$\begingroup$

I was reading through a paper, when I encounter p(a|s). What does this mean? It was in context with log probability but I cant find this notation anywhere.

  • 1
    The probability of event $a$, given that event $b$ occurs. For example, when rolling a die, $P(X\text{ is even}\mid X\text{ is prime})=\frac13$2017-02-28
  • 0
    Specifically $P(A\mid B) = \frac{P(A\cap B)}{P(B)}$. See [wiki](https://en.wikipedia.org/wiki/Conditional_probability).2017-02-28
  • 0
    Usually that notation is used for conditional probabilities. It is the likelihood of event $a$ happening, given that we know $s$ about it.2017-02-28
  • 0
    @Miemels, feel free to answer your own question, now that you know the answer!2017-02-28
  • 0
    I will add that $p(x\mid y)$ is too often used as shortened for $p_{X\mid Y}(x\mid y)$, especially in engineering and physics texts. $~$ In such case it usually means conditional probability *mass* function, which equals $\mathsf P(X=x\mid Y=y)$. $~$ ( Only occasionally is $p$ used to represent a probability *density* function. )2017-03-01

3 Answers 3

4

That is standard notation for conditional probability.

See here for example for further information.

0

Like everyone else said, it's just notation for conditional probability. Here is the textbook that I used last semester in my probability class. Flip to page 141 and you can read all about the conditional probability, both in discrete and continuous cases.

0

The notation P(x|y) means P(x) given event y has occurred, this notation is used in conditional probability.

There are two cases if x and y are dependent or if x and y are independent.

Case 1) P(x|y) = P(x&y)/P(y)

Case 2) P(x|y) = P(x)

  • 0
    "There are two cases if x and y are dependent or if x and y are independent" Actually there is only one case, when P(x|y) = P(x&y)/P(y).2017-02-28
  • 0
    Indeed, the "second case" is just a simplification; because independence means that $\mathsf P(x \& y)=\mathsf P(x)\cdot\mathsf P(y)$ .2017-02-28