Marginal probability while knowing conditional probability and the marginal of the conditioning

20 Views Asked by At

Let $Y|M = m$ be a discrete random variable with probability function $P(Y = y|M = m) = \frac{e^{-m}m^y}{y!}$ for $y \in \{0, 1, ...\}$ and $M$ a continuous random variable with density function $f_M(m) = e^{-m}$ for $m > 0$. What's the marginal probability of $Y$?

I'm aware that $Y|M = m$ ~ Poisson(m) and $M$ ~ Exponential(1), but i can't go further. Maybe i'm missing a ridiculous point.

I thought i could find $P(Y = y)$ by integrating the conditional in the range of $M$, but i was wrong. Can someone help me, please? (Sorry about my english, it's not my first language).

1

There are 1 best solutions below

1
On BEST ANSWER

I thought i could find P(Y=y) by integrating the conditional in the range of M , but i was wrong.

Don't integrate just the conditional, but the joint measure.

By the Law of Total Probability:

$$\begin{align}\mathsf P(Y{\,=\,}y) &= \int_\Bbb R \mathsf P(Y{\,=\,}y\mid M{\,=\,}m) f_M(m)\,\mathrm d m\\ &= \dfrac{1}{y!}\mathbf 1_{y\in\Bbb N}\int_0^\infty {\mathrm e^{-2m} m^y}\,\mathrm d m \\ &~~\vdots\end{align}$$