Derive distribution from normal distribution with unknown mean

46 Views Asked by At

I'm having trouble with this question since mean is unknown. If the mean is centered around 0, I think I can solve.

Let be with ∼(, 1). Define a set of new random variables X such that

X={1 if >0, 0 if <=0

Question is: Derive the distribution of X and show that it depends on .

1

There are 1 best solutions below

4
On

$X_i$ takes only two values $0$ and $1$. So finding the distribution of $X_i$ is just finding the probabilities that this variable takes the values $0$ and $1$.

$P(X_i=1)=P(Z_i >0)=\frac 1 {\sqrt {2\pi}}\int_0^{\infty} e^{-(x-\mu)^{2}/2} dx =\frac 1 {\sqrt {2\pi}}\int_{-\mu}^{\infty} e^{-y^{2}/2} dy=1-\Phi (-\mu)$ and $P(X_i=0)=1-P(X_i=1)=\Phi (-\mu)$ where $\Phi$ is the standard normal distribution function.