Annealing of iterative stochastic schemes

被引：30

作者：

Fang, HT

Gong, GL

Qian, MP

机构：

[1] BEIJING UNIV,CTR NONLINEAR SCI,BEIJING 100871,PEOPLES R CHINA

[2] TSING HUA UNIV,DEPT APPL MATH,BEIJING 100084,PEOPLES R CHINA

来源：

SIAM JOURNAL ON CONTROL AND OPTIMIZATION | 1997年 / 35卷 / 06期

关键词：

global minimum; iterative stochastic schemes; large deviation; potential function; simulated annealing;

D O I：

10.1137/S0363012995293670

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Annealing of general iterative stochastic schemes is studied by using Wentzell's large deviation theory. The convergence of this algorithm is shown under some conditions. The closed form of the critical constant is specified in terms of its potential function of the mean model.

引用

页码：1886 / 1907

页数：22

共 25 条

[1]

[Anonymous], SCIENCE

[2]

Benveniste A, 1990, Adaptive algorithms and stochastic approximations

[3] ROUGH LARGE DEVIATION ESTIMATES FOR SIMULATED ANNEALING - APPLICATION TO EXPONENTIAL SCHEDULES [J].