The Mixing Time Evolution of Glauber Dynamics for the Mean-Field Ising Model

被引:40
作者
Ding, Jian [1 ]
Lubetzky, Eyal [2 ]
Peres, Yuval [2 ]
机构
[1] Univ Calif Berkeley, Dept Stat, Berkeley, CA 94720 USA
[2] Microsoft Res, Redmond, WA 98052 USA
基金
美国国家科学基金会;
关键词
LIMIT-THEOREMS; SUMS;
D O I
10.1007/s00220-009-0781-9
中图分类号
O4 [物理学];
学科分类号
0702 ;
摘要
We consider Glauber dynamics for the Ising model on the complete graph on n vertices, known as the Curie-Weiss model. It is well-known that the mixing-time in the high temperature regime (beta < 1) has order n log n, whereas the mixing-time in the case beta > 1 is exponential in n. Recently, Levin, Luczak and Peres proved that for any fixed beta < 1 there is cutoff at time 1 /2(1-beta) n log n with a window of order n, whereas the mixing-time at the critical temperature beta = 1 is Theta (n(3/2)). It is natural to ask how the mixing-time transitions from Theta(n log n) to Theta (n(3/2)) and finally to exp (Theta(n)). That is, how does the mixing-time behave when beta = beta( n) is allowed to tend to 1 as n -> infinity. In this work, we obtain a complete characterization of the mixing-time of the dynamics as a function of the temperature, as it approaches its critical point beta(c) = 1. In particular, we find a scaling window of order 1/root n around the critical temperature. In the high temperature regime, beta = 1 - delta for some 0 < delta < 1 so that delta(2)n -> infinity with n, the mixing-time has order (n/delta) log( d2n), and exhibits cutoff with constant 1 2 and window size n/delta. In the critical window, beta = 1 +/- delta, where delta(2)n is O(1), there is no cutoff, and the mixing-time has order n(3/2). At low temperature, beta = 1+delta ford > 0 with delta(2)n -> infinity and delta = o(1), there is no cutoff, and the mixing time has order n/delta exp ((3/4 + o(1))delta(2)n).
引用
收藏
页码:725 / 764
页数:40
相关论文
共 13 条
[1]  
[Anonymous], 1987, PERCOLATION THEORY E, DOI DOI 10.1007/978-1-4613-8734-3_1
[2]  
[Anonymous], 2009, American Mathematical Soc.
[3]  
[Anonymous], REVERSIBLE MAR UNPUB
[4]   Path coupling: A technique for proving rapid mixing in Markov chains [J].
Bubley, R ;
Dyer, M .
38TH ANNUAL SYMPOSIUM ON FOUNDATIONS OF COMPUTER SCIENCE, PROCEEDINGS, 1997, :223-231
[5]  
CHEN MF, 1998, LECT NOTES STAT, V128, P123
[6]  
DING J, 2008, CENSORED GLAUBER DYN
[7]   Total variation cutoff in birth-and-death chains [J].
Ding, Jian ;
Lubetzky, Eyal ;
Peres, Yuval .
PROBABILITY THEORY AND RELATED FIELDS, 2010, 146 (1-2) :61-85
[8]  
Ellis Richard S., 1985, GRUNDLEHREN MATH WIS, V271
[9]   LIMIT-THEOREMS FOR SUMS OF DEPENDENT RANDOM-VARIABLES OCCURRING IN STATISTICAL-MECHANICS .2. CONDITIONING, MULTIPLE PHASES, AND METASTABILITY [J].
ELLIS, RS ;
NEWMAN, CM ;
ROSEN, JS .
ZEITSCHRIFT FUR WAHRSCHEINLICHKEITSTHEORIE UND VERWANDTE GEBIETE, 1980, 51 (02) :153-169
[10]   LIMIT-THEOREMS FOR SUMS OF DEPENDENT RANDOM-VARIABLES OCCURRING IN STATISTICAL-MECHANICS [J].
ELLIS, RS ;
NEWMAN, CM .
ZEITSCHRIFT FUR WAHRSCHEINLICHKEITSTHEORIE UND VERWANDTE GEBIETE, 1978, 44 (02) :117-139