随机方差缩减梯度Langevin动力学由随机时滞微分方程的逼近
概率论
2021-12-21 v2 最优化与控制
摘要
本文研究在Wasserstein-1距离下用随机时滞微分方程对随机方差缩减梯度Langevin动力学进行弱逼近，并得到一致误差界。我们的方法基于精细的Lindeberg原理与Malliavin微积分。
引用
@article{arxiv.2106.04357,
title = {Approximation to stochastic variance reduced gradient Langevin dynamics by stochastic delay differential equations},
author = {Peng Chen and Jianya Lu and Lihu Xu},
journal= {arXiv preprint arXiv:2106.04357},
year = {2021}
}
备注
We update the assumption and the convergence rate, which are both better than the original version