In this paper,we propose an accelerated stochastic variance reduction gradient method with a trust-region-like framework,referred as the NMSVRG-TR method.Based on NMSVRG,we incorporate a Katyusha-like acceleration ste...In this paper,we propose an accelerated stochastic variance reduction gradient method with a trust-region-like framework,referred as the NMSVRG-TR method.Based on NMSVRG,we incorporate a Katyusha-like acceleration step into the stochastic trust region scheme,which improves the convergence rate of the SVRG methods.Under appropriate assumptions,the linear convergence of the algorithm is provided for strongly convex objective functions.Numerical experiment results show that our algorithm is generally superior to some existing stochastic gradient methods.展开更多
本文研究了优化问题中的一类复合优化问题。对于凸非光滑的目标函数,在临近牛顿算法的基础上,引入方差缩减的方法,提出了一种新的一一基于方差缩减的随机牛顿算法(SN V R),并进行了收敛性分析。与ProxSGD,ProxGD,ProxSV RG方法相比,SN ...本文研究了优化问题中的一类复合优化问题。对于凸非光滑的目标函数,在临近牛顿算法的基础上,引入方差缩减的方法,提出了一种新的一一基于方差缩减的随机牛顿算法(SN V R),并进行了收敛性分析。与ProxSGD,ProxGD,ProxSV RG方法相比,SN V R有更快的收敛速度。展开更多
文摘In this paper,we propose an accelerated stochastic variance reduction gradient method with a trust-region-like framework,referred as the NMSVRG-TR method.Based on NMSVRG,we incorporate a Katyusha-like acceleration step into the stochastic trust region scheme,which improves the convergence rate of the SVRG methods.Under appropriate assumptions,the linear convergence of the algorithm is provided for strongly convex objective functions.Numerical experiment results show that our algorithm is generally superior to some existing stochastic gradient methods.
