arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2103.16764cs.LGmath.OC
收藏

神经网络训练的阻尼牛顿随机梯度下降法研究

Research of Damped Newton Stochastic Gradient Descent Method for Neural Network Training

Beihang University(北京航空航天大学)

Jingcheng Zhou, Wei Wei, Zhiming Zheng

摘要:

随机梯度下降法(SGD)等一阶方法是近年来训练深度神经网络(DNN)的主流优化方法,而二阶方法由于在获取高阶信息时计算代价过高而很少使用。本文针对均方误差(MSE)回归问题和交叉熵损失(CEL)分类问题,提出了阻尼牛顿随机梯度下降(DN-SGD)方法和随机梯度下降阻尼牛顿(SGD-DN)方法来训练DNN。与其他二阶方法估计所有参数的海森矩阵不同,该方法只精确计算一小部分参数,大大降低了计算量,使得学习过程的收敛速度和精度都比SGD方法快得多。在真实数据集上进行了几个数值实验,以验证我们的方法对回归和分类问题的有效性。

英文摘要:

First-order methods like stochastic gradient descent(SGD) are recently the popular optimization method to train deep neural networks (DNNs), but second-order methods are scarcely used because of the overpriced computing cost in getting the high-order information. In this paper, we propose the Damped Newton Stochastic Gradient Descent(DN-SGD) method and Stochastic Gradient Descent Damped Newton(SGD-DN) method to train DNNs for regression problems with Mean Square Error(MSE) and classification problems with Cross-Entropy Loss(CEL), which is inspired by a proved fact that the hessian matrix of last layer of DNNs is always semi-definite. Different from other second-order methods to estimate the hessian matrix of all parameters, our methods just accurately compute a small part of the parameters, which greatly reduces the computational cost and makes convergence of the learning process much faster and more accurate than SGD. Several numerical experiments on real datesets are performed to verify the effectiveness of our methods for regression and classification problems.

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。

↑