On the contraction method with degenerate limit equation

• 成果类型: Article
• 署名作者: Neininger, R; Rüschendorf, L
• 署名单位: Goethe University Frankfurt; University of Freiburg
• 刊物名称: ANNALS OF PROBABILITY
• ISSN/ISSBN: 0091-1798
• DOI: 10.1214/009117904000000171
• 发表日期: 2004
• 页码: 2838-2856
• 关键词: independent random-variables; recursive algorithms; distributions; THEOREM; quicksort; METRICS; sums

摘要

A class of random recursive sequences (Y-n) with slowly varying variances as arising for parameters of random trees or recursive algorithms X leads after normalizations to degenerate limit equations of the form X =(L) X. For nondegenerate limit equations the contraction method is a main tool to establish convergence of the scaled sequence to the unique solution of the limit equation. In this paper we develop an extension of the contraction method which allows us to derive limit theorems for parameters of algorithms and data structures with degenerate limit equation. In particular, we establish some new tools and a general convergence scheme, which transfers information on mean and variance into a central limit law (with normal limit). We also obtain a convergence rate result. For the proof we use selfdecomposability properties of the limit normal distribution which allow us to mimic the recursive sequence by an accompanying sequence in normal variables.