摘要
针对机器学习中一类有限光滑凸函数和的最小化问题,将自适应步长与SARAH++算法结合,提出了一种改进的算法SARAH++AS.然后在强凸的假设下证明了它的收敛性.最后从实验结果分析来看,相比于使用固定步长的SARAH++算法,新算法的收敛速度更快,不受初始步长选取的影响.新算法对初始步长的选择是有效的.
Aiming at the minimization problem of finite smooth convex function sum in machine learning,an improved algorithm SARAH++AS is proposed by combining adaptive step size with SARAH++algorithm.Then its convergence is proved under the strong convexity hypothesis.Finally,from the analysis of experimental results,compared with the fixed step size SARAH++algorithm,the convergence speed of the new algorithm is faster and less affected.The effect of the initial step selection.The new algorithm is effective for selecting the initial step size.
作者
李晓桐
王福胜
乔晓云
LI Xiaotong;WANG Fusheng;QIAO Xiaoyun(School of Mathematics and Statistics,Taiyuan Normal University;Basic Courses Department,Shanxi Vocational University of Engineering Science and Technology,Jinzhong Shanxi 030619,China)
出处
《太原师范学院学报(自然科学版)》
2023年第4期25-30,共6页
Journal of Taiyuan Normal University:Natural Science Edition
基金
山西省基础研究计划(自由探索类)面上项目(202103021224303)。
关键词
自适应步长
随机递归
梯度下降
adaptive step size
random recursion
gradient descent