section 2.2.7 of "Machine Learning: A Probabilistic Perspective by Kevin Patrick Murphy" gives this formula for expected value.

${\displaystyle \operatorname {E} [X]=\int_{\mathcal{X} }xf(x)\,dx.}$

There is a $\mathcal{X}$ at the bottom of integral symbol.

Assume $\mathcal{X}$ is a continuous real value random variable, since the book is discussing Machine Learning.

What does that $\mathcal{X}$ mean here? Does it mean "the domain of definition", "support" or something else?

  • $\begingroup$ @Masacroso It's not the domain of $X$. In fact, $\mathbb X$ should be $\mathbb R$ if $X$ is a real continuous random variable. $\endgroup$ – amsmath Oct 14 '19 at 4:26
  • $\begingroup$ @amsmath you are right, my bad. Here $\Bbb X$ is the domain of $f$, not of $X$. $\endgroup$ – Masacroso Oct 14 '19 at 4:27
  • $\begingroup$ See here: en.wikipedia.org/wiki/… $\endgroup$ – amsmath Oct 14 '19 at 4:31
  • $\begingroup$ @amsmath Thank you! In your link, "the uniform distribution on the interval [0, ½] has probability density f(x) = 2 for 0 ≤ x ≤ ½ and f(x) = 0 elsewhere", is the interval [0, ½] the support of that uniform distribution, right? $\endgroup$ – JJJohn Oct 14 '19 at 4:45
  • $\begingroup$ @baojieqh Exactly! $\endgroup$ – amsmath Oct 14 '19 at 4:55

$\Bbb X$ (or $\mathcal X$ after the edit) should indeed be the support for $X$, the domain over which the probability density function, $f$, returns non-zero values.

If $X$ is a real-valued random variable, then $\Bbb X\subseteq \Bbb R$.


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.