The growth of iterates of multivariate generating functions期刊界 All Journals 搜尽天下杂志传播学术成果专业期刊搜索期刊信息化学术搜索

按检索

The growth of iterates of multivariate generating functions

Authors:	J D Biggins

Institution:	Department of Probability and Statistics, The University of Sheffield, Sheffield, S3 7RH, United Kingdom

Abstract:	The vector-valued function $m(\theta)$ of a -vector $\theta$ has components $m_1(\theta), m_2(\theta), \dots, m_p(\theta)$ . For each , $\exp(m_i(-\theta))$ is the (multivariate) Laplace transform of a discrete measure concentrated on $0,\infty)^p$ with only a finite number of atoms. The main objective is to give conditions for the functional iterates $m^{(n)}$ of to grow like $\rho^n$ for a suitable $\rho>1$ . The initial stimulus was provided by results of Miller and O'Sullivan (1992) on enumeration issues in `context free languages', results which can be improved using the theory developed here. The theory also allows certain results in Jones (2004) on multitype branching to be proved under significantly weaker conditions.

Keywords:	Generating functions iterates maximum growth rate non-negative matrices functional iteration multitype branching processes enumeration

	点击此处可从《Transactions of the American Mathematical Society》浏览原始摘要信息
	点击此处可从《Transactions of the American Mathematical Society》下载免费的PDF全文