1

I saw these in a machine learning notes. and like to know the meaning of them? also the formula of them?

shadowtalker
  • 11,395
  • 3
  • 49
  • 109
WDR
  • 131
  • 1
  • 4

1 Answers1

1

welcome to SO! They are simply a notation for conditional probabilities. You often see them in Bayesian statistics.

How to interpret it:

For instance, your example $p(y|x;\theta)$ means what is the probability that you observe $y$, when $x$ and $\theta$ 'holds'. Usually, $x$ stands for some data and $\theta$ parameters of distribution, or a model. It can be the probability of observing something given the dataset $x$ and estimated linear regression coefficients $\theta = (\beta_0,\beta_1)$.

Have a look for instance at Wiki page

Jan Sila
  • 305
  • 1
  • 10
  • Thanks, but it didn't say about p(x;y) , p(x|y,z) and p(x|y;z) – WDR Aug 22 '17 at 12:36
  • that is just the notation of conditional probability, if not stated otherwise – Jan Sila Aug 22 '17 at 12:40
  • This is being automatically flagged as low quality, probably because it is so short. At present it is more of a comment than an answer by our standards. Can you expand on it? You can also turn it into a comment. – gung - Reinstate Monica Aug 22 '17 at 13:52