English

On corrected Poisson approximations for sums of independent indicators

Probability 2023-05-08 v2

Abstract

Let Sn=I1++InS_n=I_1+\cdots+I_n be a sum of independent indicators IiI_i, with pi=Pr(Ii=1)=1Pr(Ii=0)p_i=\Pr(I_i=1)=1-\Pr(I_i=0), i=1,,ni=1,\ldots,n. It is well-known that the total variation distance between SnS_n and ZλZ_\lambda, where ZλZ_\lambda has a Poisson distribution with mean λ=i=1npi\lambda=\sum_{i=1}^n p_i, is typically of order i=1npi2\sum_{i=1}^n p_i^2. In the present work we propose a class of corrected Poisson approximations, which enable the second order factorial moment distance (and hence, the total variation distance) to be bounded above by a constant multiple of i=1npi3\sum_{i=1}^n p_i^3 and i=1npi4\sum_{i=1}^n p_i^4, hence improving the order of approximation.

Keywords

Cite

@article{arxiv.2304.10314,
  title  = {On corrected Poisson approximations for sums of independent indicators},
  author = {Nickos Papadatos},
  journal= {arXiv preprint arXiv:2304.10314},
  year   = {2023}
}

Comments

15 pages, 1 Table