Randomized sequential importance sampling for estimating the number of perfect matchings in bipartite graphs

被引:2
|
作者
Diaconis, Persi [1 ]
Kolesnik, Brett [2 ]
机构
[1] Stanford Univ, Dept Math & Stat, Stanford, CA 94305 USA
[2] Univ Calif Berkeley, Dept Stat, Berkeley, CA 94720 USA
基金
加拿大自然科学与工程研究理事会;
关键词
Asymptotic normality; Distributional recursion; Importance sampling; Monte Carlo methods; Perfect matchings; Randomized algorithms; RECURRENCE; PERMANENT;
D O I
10.1016/j.aam.2021.102247
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We introduce and study randomized sequential importance sampling algorithms for estimating the number of perfect matchings in bipartite graphs. In analyzing their performance, we establish various non-standard central limit theorems. We expect our methods to be useful for other applied problems. (c) 2021 Published by Elsevier Inc.
引用
收藏
页数:41
相关论文
共 50 条