arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.02882stat.MLcs.DCcs.LGmath.OC
收藏

非凸优化的按比例放大随机梯度下降算法

Scaling up Stochastic Gradient Descent for Non-convex Optimisation

Bournemouth University(伯恩茅斯大学);Warwick University(华威大学)

Saad Mohamad, Hamad Alamri, Abdelhamid Bouchachia

摘要:

随机梯度下降(SGD)是一种被广泛采用的优化可微目标函数的迭代方法。在本文中,我们提出并讨论了一种新的方法来扩大SGD在涉及非凸函数和大型数据集的应用。我们解决了使用共享和分布式内存时出现的瓶颈问题。通常,前者受有限的计算资源和带宽的限制,而后者则受到通信开销的影响。我们提出了一个统一的分布式和并行实现的SGD(命名为DPSGD),依赖于异步分布和无锁并行。通过将两种策略结合到一个统一的框架中,DPSGD能够在本地计算和通信之间取得更好的平衡。DPSGD的收敛性能研究的非凸问题,如统计建模和机器学习中出现的。我们的理论分析表明,DPSGD导致的速度方面的核心和工人的数量,同时保证$O(1/\sqrt{T})$的渐近收敛速度给定的核心的数量是由$T^{1/4}$和工人的数量是由$T^{1/2}$,其中$T$是迭代次数的限制。DPSGD可以实现的潜在收益在随机变分推断问题(潜在狄利克雷分配)和深度强化学习(DRL)问题(优势演员评论家-A2 C)上进行了经验证明,从而产生了两种算法:DPSVI和HSA 2C。实证结果验证了我们的理论研究结果。进行比较研究,以显示所提出的DPSGD对国家的最先进的DRL算法的性能。

英文摘要:

Stochastic gradient descent (SGD) is a widely adopted iterative method for optimizing differentiable objective functions. In this paper, we propose and discuss a novel approach to scale up SGD in applications involving non-convex functions and large datasets. We address the bottleneck problem arising when using both shared and distributed memory. Typically, the former is bounded by limited computation resources and bandwidth whereas the latter suffers from communication overheads. We propose a unified distributed and parallel implementation of SGD (named DPSGD) that relies on both asynchronous distribution and lock-free parallelism. By combining two strategies into a unified framework, DPSGD is able to strike a better trade-off between local computation and communication. The convergence properties of DPSGD are studied for non-convex problems such as those arising in statistical modelling and machine learning. Our theoretical analysis shows that DPSGD leads to speed-up with respect to the number of cores and number of workers while guaranteeing an asymptotic convergence rate of $O(1/\sqrt{T})$ given that the number of cores is bounded by $T^{1/4}$ and the number of workers is bounded by $T^{1/2}$ where $T$ is the number of iterations. The potential gains that can be achieved by DPSGD are demonstrated empirically on a stochastic variational inference problem (Latent Dirichlet Allocation) and on a deep reinforcement learning (DRL) problem (advantage actor critic - A2C) resulting in two algorithms: DPSVI and HSA2C. Empirical results validate our theoretical findings. Comparative studies are conducted to show the performance of the proposed DPSGD against the state-of-the-art DRL algorithms.

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。

↑