中文

progeny 均值无穷的枝杈随机游动:双尾之谈

概率论 2022-07-05 v4

摘要

我们在底层 Galton-Watson 树具有无穷 progeny 均值的假设下研究枝杈随机游动的极值。假设位移要么是正则变化的,要么具有更轻的尾部。在正则变化情形下,证明了归一化极值的点过程序列收敛到 Poisson 随机测度。当位移的尾部行为形如 exp(K(x))\exp(-K(x)) 时,我们研究第 nn 代最右粒子缩放位置的渐近性,其中 KK 要么是指数为 r>0r>0 的正则变化函数,要么 KK 具有指数增长。我们在所有情形下确定了极大值的确切缩放,并证明了当 r>1r>1 时非平凡极限的存在性。

关键词

引用

@article{arxiv.1909.08948,
  title  = {Branching random walk with infinite progeny mean: a tale of two tails},
  author = {Souvik Ray and Rajat Subhra Hazra and Parthanil Roy and Philippe Soulier},
  journal= {arXiv preprint arXiv:1909.08948},
  year   = {2022}
}

备注

33 pages. Improved version and contains many new results. Section 5 and section 6 added on very rapidly varying tails and cloudspeed respectively. The proofs are streamlined and many new arguments added