3
$\begingroup$

I am trying to understand Neyman Pearson Lemma's proff from Rice's book. The lemma is intuitive, however I am not able to understand the reasoning for the first inequality in the proof. I highlighted their reasoning (just below the inequality), but not able to follow it.enter image description here

I also believe there is a typo where 0 should be 1 (highlighted in pink).

$\endgroup$

1 Answer 1

4
$\begingroup$

First note that by definition of $d(x)$ we have $f_0(x)/f_A(x)\ge c$ when $d(x)=0$ and $f_0(x)/f_A(x)\lt c$ when $d(x)=1$. These inequalities are equivalent to $c f_A(x)-f_0(x)\le 0$ and $c f_A(x)-f_0(x)\gt 0$ respectively.

Now consider the case where $d(x)=0$. Then the right hand side of the inequality to be proven is zero and the left hand side is non-positive. The latter follows from the fact that $d^*(x)\ge 0$ and also because we have seen that $c f_A(x)-f_0(x)\le 0$ when $d(x)=0$. So the inequality holds when $d(x)=0$.

Next consider the case where $d(x)=1$ and $d*(x)=0$. In this case, the right hand side of the inequality to be proven is $c f_A(x)-f_0(x)$, which must be positive since $d(x)=1$. Since the left hand side is zero in this case, again the inequality holds.

The last possibility is the case where both $d(x)=1$ and $d^*(x)=1$. In this case both sides of the inequality to be proven are equal to $c f_A(x)-f_0(x)$, and so again the inequality holds.

QED

BTW, you are correct about the typo.

$\endgroup$
1
  • $\begingroup$ Awesome ... Thank you $\endgroup$ Commented Mar 1 at 17:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.