Knowledge Commons of Institute of Automation,CAS
Stochastic Conjugate Gradient Algorithm With Variance Reduction | |
Jin, Xiao-Bo1; Zhang, Xu-Yao2; Huang, Kaizhu3; Geng, Guang-Gang4 | |
发表期刊 | IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS |
ISSN | 2162-237X |
2019-05-01 | |
卷号 | 30期号:5页码:1360-1369 |
通讯作者 | Geng, Guang-Gang(gengguanggang@cnnic.cn) |
摘要 | Conjugate gradient (CG) methods are a class of important methods for solving linear equations and nonlinear optimization problems. In this paper, we propose a new stochastic CG algorithm with variance reduction(1) and we prove its linear convergence with the Fletcher and Reeves method for strongly convex and smooth functions. We experimentally demonstrate that the CG with variance reduction algorithm converges faster than its counterparts for four learning models, which may be convex, nonconvex or nonsmooth. In addition, its area under the curve performance on six large-scale data sets is comparable to that of the LIBLINEAR solver for the L2-regularized L2-loss but with a significant improvement in computational efficiency. |
关键词 | Computational efficiency covariance reduction empirical risk minimization (ERM) linear convergence stochastic conjugate gradient (CG) |
DOI | 10.1109/TNNLS.2018.2868835 |
关键词[WOS] | GLOBAL CONVERGENCE |
收录类别 | SCI |
语种 | 英语 |
资助项目 | Fundamental Research Funds for the Henan Provincial Colleges and Universities in Henan University of Technology[2016RCJH06] ; National Natural Science Foundation of China[61103138] ; National Natural Science Foundation of China[61602154] ; National Natural Science Foundation of China[61375039] ; National Natural Science Foundation of China[61473236] ; National Natural Science Foundation of China[U1804326] ; National Key Research & Development Program[2016YFD0400104-5] ; National Basic Research Program of China[2012CB316301] ; Fundamental Research Funds for the Henan Provincial Colleges and Universities in Henan University of Technology[2016RCJH06] ; National Natural Science Foundation of China[61103138] ; National Natural Science Foundation of China[61602154] ; National Natural Science Foundation of China[61375039] ; National Natural Science Foundation of China[61473236] ; National Natural Science Foundation of China[U1804326] ; National Key Research & Development Program[2016YFD0400104-5] ; National Basic Research Program of China[2012CB316301] |
项目资助者 | Fundamental Research Funds for the Henan Provincial Colleges and Universities in Henan University of Technology ; National Natural Science Foundation of China ; National Key Research & Development Program ; National Basic Research Program of China |
WOS研究方向 | Computer Science ; Engineering |
WOS类目 | Computer Science, Artificial Intelligence ; Computer Science, Hardware & Architecture ; Computer Science, Theory & Methods ; Engineering, Electrical & Electronic |
WOS记录号 | WOS:000466192100007 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
引用统计 | |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/24224 |
专题 | 多模态人工智能系统全国重点实验室_模式分析与学习 |
通讯作者 | Geng, Guang-Gang |
作者单位 | 1.Henan Univ Technol, Dept Informat Sci & Engn, Zhengzhou 450001, Henan, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 3.Xian Jiaotong Liverpool Univ, Dept Elect & Elect Engn, Suzhou 215123, Peoples R China 4.Chinese Acad Sci, Comp Network Informat Ctr, Beijing 100190, Peoples R China |
推荐引用方式 GB/T 7714 | Jin, Xiao-Bo,Zhang, Xu-Yao,Huang, Kaizhu,et al. Stochastic Conjugate Gradient Algorithm With Variance Reduction[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,2019,30(5):1360-1369. |
APA | Jin, Xiao-Bo,Zhang, Xu-Yao,Huang, Kaizhu,&Geng, Guang-Gang.(2019).Stochastic Conjugate Gradient Algorithm With Variance Reduction.IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,30(5),1360-1369. |
MLA | Jin, Xiao-Bo,et al."Stochastic Conjugate Gradient Algorithm With Variance Reduction".IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 30.5(2019):1360-1369. |
条目包含的文件 | 条目无相关文件。 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论