Risk Function as Expected Value of Loss Function Evaluation

419 Views Asked by At

https://i.stack.imgur.com/BZauH.png

I'm interested in part (g), specifically. I know that the risk function is the expected value of the loss function, but I'm confused as to how to evaluate it. When I replace $\theta$ with $\tilde \theta$, I get an expression which is pretty difficult to evaluate the expected value of... so perhaps I'm not going about this the right way. How would you calculate this risk function?

1

There are 1 best solutions below

0
On

Find risk function for $\bar X$. $$\mathbb E[X_1]=\frac{1-p}{p}=\theta, \ \text{Var}[X_1]= \frac{1-p}{p^2}=\theta^2+\theta. $$ Then $$\mathbb E[\bar X]=\theta, \ \text{Var}[\bar X]= \frac{\theta^2+\theta}n.$$ Risk runction for $\bar X$ is $$ \mathbb E\left[\frac{(\bar X-\theta)^2}{\theta^2+\theta}\right] = \frac{\text{Var}[\bar X]}{\theta^2+\theta} = \frac1n. $$ Find risk function for $\tilde \theta=\frac{n+1}{n}\bar X$. $$ \mathbb E\left[\frac{\left(\frac{n+1}{n}\bar X-\theta\right)^2}{\theta^2+\theta}\right] =\frac{\mathbb E\left[\left(\frac{n+1}{n}\bar X-\theta\right)^2\right]}{\theta^2+\theta}=\frac{\text{Var}\left[\frac{n+1}{n}\bar X-\theta\right]+\left(\mathbb E\left[\frac{n+1}{n}\bar X-\theta\right]\right)^2}{\theta^2+\theta} = \frac{\frac{(n+1)^2}{n^2}\text{Var}\left[\bar X\right]+\frac{\theta^2}{n^2}}{\theta^2+\theta} =\frac{(n+1)^2}{n^3}+\frac{\theta}{n^2(\theta+1)} $$