0
$\begingroup$

I want to verify whether a given conditional probability function is valid or not.

$\mathsf P(y\mid x)=\begin{cases}c\, e^{(-y/x)} & : y\geqslant 0, x>0;\\ 0 & :\text{otherwise}\end{cases}$

How should I perform my integration, there are no marginals and no joint probability function given. Can I perform with respect to $d(y/x)$? or should I proceed using a double integral?

thanks in advance, phani tej

$\endgroup$
0
$\begingroup$

Assuming $\mathsf P$ is meant to be the conditional probability density function (pdf), you should integrate $\int_0^\infty \mathsf P(y\mid x) \operatorname d y$, which should equal $1$ iff valid.


Assuming $\mathsf P$ is meant to be the conditional cumulative probability function (CDF), then validity requires: $$\forall x\in(0,\infty):[\mathsf P(0\mid x)=0\bigwedge \lim\limits_{y\to\infty}\mathsf P(y\mid x)=1\bigwedge \forall y\in[0,\infty):\frac{\mathrm d \mathsf P(y\mid x)}{\mathrm d y}\geqslant 0]$$

$\endgroup$

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.