arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.13093cs.LGmath.OCphysics.data-an
收藏

第二类鞍点与随机梯度下降的概率稳定性

Type-II Saddles and Probabilistic Stability of Stochastic Gradient Descent

Massachusetts Institute of Technology(麻省理工学院);NTT Research(NTT研究院);Université Paris Cité(巴黎西岱大学);RIKEN Center for Emergent Matter Science (CEMS)(理化学研究所新兴物质科学中心);The University of Tokyo(东京大学)

Liu Ziyin, Botao Li, Tomer Galanti, Masahito Ueda

AI总结:

本文将神经网络鞍点分为两类,指出第二类鞍点因梯度噪声消失而难以逃离,并用随机矩阵乘积、概率稳定性与Lyapunov指数刻画SGD动力学,划分出四种相。

摘要:

刻画并理解随机梯度下降(SGD)在鞍点附近的动力学仍然是一个尚未解决的问题。我们首先表明,神经网络中的鞍点可以分为两类,其中第二类鞍点尤其难以逃离,因为梯度噪声在鞍点处会消失。因此,SGD在这些鞍点附近的动力学在主导阶上由随机矩阵乘积过程描述,于是利用概率稳定性概念及相关的Lyapunov指数来研究SGD在这些鞍点附近的动力学是自然的。在理论上,我们将SGD动力学研究与遍历理论中的著名概念联系起来,并借此表明:鞍点对SGD而言既可能是吸引的,也可能是排斥的;根据鞍点附近梯度中的信噪比,其动力学可被划分为四个不同的相。

英文摘要:

Characterizing and understanding the dynamics of stochastic gradient descent (SGD) around saddle points remains an open problem. We first show that saddle points in neural networks can be divided into two types, among which the Type-II saddles are especially difficult to escape from because the gradient noise vanishes at the saddle. The dynamics of SGD around these saddles are thus to leading order described by a random matrix product process, and it is thus natural to study the dynamics of SGD around these saddles using the notion of probabilistic stability and the related Lyapunov exponent. Theoretically, we link the study of SGD dynamics to well-known concepts in ergodic theory, which we leverage to show that saddle points can be either attractive or repulsive for SGD, and its dynamics can be classified into four different phases, depending on the signal-to-noise ratio in the gradient close to the saddle.

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。

↑