0
$\begingroup$

I am reading this tuorial on constrastive divergence and at the end of page 2 they define the expectation given the data. I am unclear on what this means? How is this different from regular expected value?

$\endgroup$
  • $\begingroup$ Maybe they just mean the empirical mean? $\endgroup$ – kjetil b halvorsen Dec 16 '16 at 12:21
  • $\begingroup$ of the sample ? $\endgroup$ – user2879934 Dec 16 '16 at 13:06
  • $\begingroup$ It would be helpful to put the relevant equations in the question. In general "expected value given the data" would refer to a combination of this and this. $\endgroup$ – GeoMatt22 Dec 16 '16 at 14:25
  • $\begingroup$ If I have found the part you mean, you misquote what it says there. It doesn't say "given the data" it says "given the data distribution". The distinction is important. $\endgroup$ – Glen_b -Reinstate Monica Dec 16 '16 at 23:33
  • $\begingroup$ What is the distinction $\endgroup$ – user2879934 Dec 17 '16 at 0:47
0
$\begingroup$

In that context, it is still referring to a population mean.

The distribution F(X) in the question has population parameter vector $\Theta$. It's just a nonstandard notation.

$\endgroup$

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.