Let $X_1, X_2, \cdots$ be i.i.d. random variables with $E(X_1) = 0, E(X_1^2) = \sigma^2 >0, E(|X_1|^3) = \rho < \infty$. Let $Y_n = \frac{1}{n} \sum_{i=1}^n X_i$ and let us note $F_n$ (resp. $\Phi$) the cumulative distribution function of $\frac{Y_n \sqrt{n}}{\sigma}$ (resp. of the standard normal distribution). Then, Berry Esseen theorem states that there exists a positive constant $C$ such that for all $x$ and $n$, $$|F_n(x)-\Phi(x)| \leq \frac{C \rho}{\sigma^3 \sqrt{n}}.$$

Are there known conditions on the distribution of $X_1$ that allow to derive a similar statement for probability density functions instead of cumulative distribution functions?

  • 2
    $\begingroup$ Sure. Look, e.g., at the supplements to chapter VII in the classical book "Sums of independent random variables" by Petrov. $\endgroup$ – Yvan Velenik Sep 19 '11 at 9:02
  • $\begingroup$ ok, thanks! I just need to find a copy of the book... $\endgroup$ – Anthony Leverrier Sep 19 '11 at 9:16
  • 1
    $\begingroup$ Hum. If you search for "Sums of independent random variables" Petrov on google.com, you should find what you need on the first page of results ;) ... $\endgroup$ – Yvan Velenik Sep 19 '11 at 9:24

The magic words are "local limit theorem".

  • $\begingroup$ Are there local limit theorems with explicit constants like Berry Esseen? I have "Normal Approximation and Asymptotic Expansions" by Bhattacharya and Rao, and all the local limit theorems seem to be assymptotic or with unspecified constants. $\endgroup$ – Thomas Dybdahl Ahle Oct 10 '19 at 13:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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