#zettel #statistics-and-probability
sourced from: [[Cramer-Rao Lower Bound]]
# How to find the [[Cramer-Rao Lower Bound]]
1. Write the log likelihood function (LLF) as a function of $\theta$
- $\text{ln} \hspace{0.2em} p(\mathbf{x}; \theta)$
2. Fix $\mathbf{x}$ and take the second partial of LLF:
- $\partial^2 \text{ln} \hspace{0.2em} p(\mathbf{x}; \theta) / \partial \theta^2$
3. If the result still depends on $\mathbf{x}$,
- Fix $\theta$ and take [[Expected Value]] w.r.t. $\mathbf{x}$
- Otherwise skip this step.
4. The result may still depend on $\theta$:
- If so, just evaluate this function at each specific value of $\theta$ that is desired.
5. Negate and form reciprocal.