Accelerating Stochastic Composition Optimization
Accelerating Stochastic Composition Optimization
复制标题
DOI：
--
复制
发表时间：
2016-07
期刊：
影响因子：
--
通讯作者：
Mengdi Wang;Ji Liu;Ethan X. Fang
中科院分区：
文献类型：
--
作者：
Mengdi Wang;Ji Liu;Ethan X. Fang
考虑随机组合优化问题，其中目标是两个期望值函数的组合。我们提出了一种新的随机一阶方法，即加速随机组合近端梯度（ASC-PG）方法，该方法基于使用两个不同时间尺度对采样预言机的查询进行更新。 ASC-PG是第一个解决随机组合问题的近端梯度方法，可以处理非平滑正则化惩罚。我们证明 ASC-PG 比最知名的算法表现出更快的收敛速度，并且在几个重要的特殊情况下实现了最佳的样本错误复杂度。我们进一步演示了ASC-PG在强化学习中的应用并进行了数值实验。
Consider the stochastic composition optimization problem where the objective is a composition of two expected-value functions. We propose a new stochastic first-order method, namely the accelerated stochastic compositional proximal gradient (ASC-PG) method, which updates based on queries to the sampling oracle using two different timescales. The ASC-PG is the first proximal gradient method for the stochastic composition problem that can deal with nonsmooth regularization penalty. We show that the ASC-PG exhibits faster convergence than the best known algorithms, and that it achieves the optimal sample-error complexity in several important special cases. We further demonstrate the application of ASC-PG to reinforcement learning and conduct numerical experiments.