English

Estimating the logarithm of characteristic function and stability parameter for symmetric stable laws

Statistics Theory 2020-08-12 v1 Statistics Theory

Abstract

Let X1,,XnX_1,\ldots,X_n be an i.i.d. sample from symmetric stable distribution with stability parameter α\alpha and scale parameter γ\gamma. Let φn\varphi_n be the empirical characteristic function. We prove an uniform large deviation inequality: given preciseness ϵ>0\epsilon>0 and probability p(0,1)p\in (0,1), there exists universal (depending on ϵ\epsilon and pp but not depending on α\alpha and γ\gamma) constant rˉ>0\bar{r}>0 so that P(supu>0:r(u)rˉr(u)r^(u)ϵ)p,P\big(\sup_{u>0:r(u)\leq \bar{r}}|r(u)-\hat{r}(u)|\geq \epsilon\big)\leq p, where r(u)=(uγ)αr(u)=(u\gamma)^{\alpha} and r^(u)=lnφn(u)\hat{r}(u)=-\ln|\varphi_n(u)|. As an applications of the result, we show how it can be used in estimation unknown stability parameter α\alpha.

Keywords

Cite

@article{arxiv.2008.04423,
  title  = {Estimating the logarithm of characteristic function and stability parameter for symmetric stable laws},
  author = {Annika Krutto and Jüri Lember},
  journal= {arXiv preprint arXiv:2008.04423},
  year   = {2020}
}
R2 v1 2026-06-23T17:45:54.167Z