2
$\begingroup$

If I have two measure space $X,Y$, with measure $\mu_X$ and $\mu_Y$, the usual way to define a measure for the combined measure space $U=X \times Y$, is the product measure, i.e., $$\mu_U = \mu_X \times \mu_Y$$

However, this implies the independence between $X$ and $Y$, in my case, I want to define a measure that has dependency on each other. The corresponding idea would be define a conditional measure, $$ \mu_{U} = \mu_{X} \mu_{Y|X} $$ It is possible? Has any one heard about anything like it?

$\endgroup$

1 Answer 1

1
$\begingroup$

Yes these are called Markov kernels. In this paper, on page 229, he also define the conditional probability using the Radon-Nikodym theorem.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .