arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.14625math.NAcs.NA
收藏

关于随机梯度下降的离散原理

On the Discrepancy Principle for Stochastic Gradient Descent

Tim Jahn, Bangti Jin

摘要:

随机梯度下降法是一种很有前途的求解大规模反问题的数值方法。然而,它的理论性质在很大程度上仍然在经典正则化理论的镜头下探索。本文研究了经典的差异原理,即最流行的\textit{a posteriori}选择规则之一,作为SGD的停止准则,证明了有限迭代终止性和当噪声趋于零时收敛概率的收敛性.理论结果与广泛的数值实验进行了补充。

英文摘要:

Stochastic gradient descent (SGD) is a promising numerical method for solving large-scale inverse problems. However, its theoretical properties remain largely underexplored in the lens of classical regularization theory. In this note, we study the classical discrepancy principle, one of the most popular \textit{a posteriori} choice rules, as the stopping criterion for SGD, and prove the finite iteration termination property and the convergence of the iterate in probability as the noise level tends to zero. The theoretical results are complemented with extensive numerical experiments.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑