The-Square-and-Add Markov Chain
The-Square-and-Add Markov Chain
复制标题
平方加马尔可夫链
DOI:
10.1007/s00283-021-10058-w
复制
发表时间:
2021
期刊:
影响因子:
--
通讯作者:
Martin Isaacs, I.
中科院分区:
文献类型:
--
作者:
Diaconis, Persi;He, Jimmy;Martin Isaacs, I.
L et us begin with a problem we cannot solve. If q is a prime power, we write Fq to denote the field with q elements, so if p is prime, Fp is the field of integers modulo p. A simple random walk (drunkard’s walk) on Fp goes from j to j þ 1 or j Ā 1 with probability 1/2. As time goes on, this converges to the uniform distribution on Fp. This means that after a long time, the probability that the random walk will be at some a 2 Fp is about 1/p. It takes about p2 steps for this convergence to kick in. This is slow: if p ž 101, then p2 ž 10 201. These informal statements are explained more carefully after Theorem 1 below. One attempt to speed things up intersperses deterministic doubling with the random Æ1 steps. If Xn denotes the position of the walk after n steps (say starting from X0 ž 0), then this new walk is
登录
查看更多内容
影响因子:
2
作者:
S. Chatterjee;P. Diaconis
通讯作者:
S. Chatterjee;P. Diaconis
影响因子:
1.4
作者:
Jimmy He
通讯作者:
Jimmy He
影响因子:
0.7
作者:
S. Lenwood;A. L. Nicholas
通讯作者:
A. L. Nicholas
影响因子:
0.7
作者:
D. A. Levin;Y. Peres;Elizabeth L. Wilmer
通讯作者:
Elizabeth L. Wilmer
影响因子:
2.3
作者:
F. Chung;P. Diaconis;R. Graham
通讯作者:
R. Graham