arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2104.05468math.OC
收藏

L-光滑函数固定步长梯度法的精确最坏收敛速度

The exact worst-case convergence rate of the gradient method with fixed step lengths for L-smooth functions

Hadi Abbaszadehpeivasti, Etienne de Klerk, Moslem Zamani

摘要:

本文研究了求解$L$ -光滑函数的固定步长梯度法(或最速下降法)的收敛速度。我们建立了一个新的收敛速度,并表明,在某些情况下,特别是当所有的步长位于区间$(0,1/L]$的界限可能是准确的。此外,我们推导出一个最佳的步长的新的界限。

英文摘要:

In this paper, we study the convergence rate of the gradient (or steepest descent) method with fixed step lengths for finding a stationary point of an $L$-smooth function. We establish a new convergence rate, and show that the bound may be exact in some cases, in particular when all step lengths lie in the interval $(0,1/L]$. In addition, we derive an optimal step length with respect to the new bound.

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。

↑