General results on the convergence of stochastic algorithms

被引：50

作者：

Delyon, B

机构：

[1] IRISA/INRIA, Campus de Beaulieu

来源：

IEEE TRANSACTIONS ON AUTOMATIC CONTROL | 1996年 / 41卷 / 09期

关键词：

D O I：

10.1109/9.536495

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

A deterministic approach is proposed for proving the convergence of stochastic algorithms of the most general form under necessary conditions on the input noise and reasonable conditions on the (nonnecessarily continuous) mean field, Emphasis is placed on the case where more than one stationary point exists, We also use this approach to prove the convergence of a stochastic algorithm with Markovian dynamics.

引用

页码：1245 / 1255

页数：11

共 20 条

[1]

[Anonymous], 1992, Stochastic Stability of Markov chains

[2]

Benveniste A, 1990, Adaptive algorithms and stochastic approximations

[3]

BRANDIERE O, IN PRESS ANN IHP

[4]

DUFLO M., 1990, Methodes Recursives Aleatoires

[5]

FIJALKOW I, UNPUB IEEE T SIGNAL

[6]

Gantmacher F R, 1966, THEORY MATRICES

[7]

GLYNN PW, IN PRESS AN PROBAB

[8]

HALL P, 1980, MARTINGALE LIMIT THE

[9]

Kato T., 1976, Grundlehren der Mathematischen Wissenschaften, V132

[10]

KRASNOVSKII, STABILITY MOTION

← 1 2 →