期刊
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS
卷 179, 期 3, 页码 838-867出版社
SPRINGER/PLENUM PUBLISHERS
DOI: 10.1007/s10957-018-1346-x
关键词
Mirror descent; Variational inequalities; Saddle-point problems; Stochastic differential equations; 90C25; 90C33; 90C47
资金
- French National Research Agency (ANR) [ANR- 16- CE33- 0004- 01]
- COST Action [CA16228]
We examine a class of stochastic mirror descent dynamics in the context of monotone variational inequalities (including Nash equilibrium and saddle-point problems). The dynamics under study are formulated as a stochastic differential equation, driven by a (single-valued) monotone operator and perturbed by a Brownian motion. The system's controllable parameters are two variable weight sequences, that, respectively, pre- and post-multiply the driver of the process. By carefully tuning these parameters, we obtain global convergence in the ergodic sense, and we estimate the average rate of convergence of the process. We also establish a large deviations principle, showing that individual trajectories exhibit exponential concentration around this average.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据