{"id":"arxiv/2610.24413","url":"https://arxiv.org/abs/2610.24413","tags":["math.ST","math.PR"],"score":0,"title":"Asymptotic Normality of the Hill Estimator in Linear Preferential Attachment","domain":"math.ST","source":"arxiv","authors":["Tiandong Wang"],"reviews":0,"subject":"math.ST","summary":"We prove asymptotic normality of the Hill estimator based on the largest $k_n$ degrees in a homogeneous linear preferential attachment graph with $n$ vertices. Each new vertex sends a fixed number $m\\ge1$ of edges sequentially to older vertices, with attachment weight $d+δ$ and fixed $δ>-m$. Let $α=m/(2m+δ)$. For every deterministic sequence $k_n\\to\\infty$ satisfying $\\sqrt{k_n}(k_n/n)^α\\to0$, the estimator, centered at $α$ and multiplied by $\\sqrt{k_n}$, converges to a centered normal law with an explicit variance strictly smaller than $α^2$. The reduction relative to the Hill variance for…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-17","reactions":{}}