arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2205.04491cs.LGmath.STstat.TH
收藏

浅层神经网络近似 stationary 点的统计保证

Statistical Guarantees for Approximate Stationary Points of Shallow Neural Networks

University of Hamburg(汉堡大学);Beijing Normal-Hong Kong Baptist University(北京师范大学-香港浸会大学联合国际学院)

Mahsa Taheri, Fang Xie, Johannes Lederer

AI总结:

本文研究了浅层神经网络近似 stationary 点的统计保证,提出方法支持神经网络无需全局优化的数学观点,并扩展至 ReLU 神经网络。

摘要:

由于神经网络的统计保证通常局限于复杂目标函数的全局极值,因此不清楚这些理论是否能够解释实际神经网络流水线的输出性能。本文的目标是将统计理论与实践联系起来。我们开发了适用于浅层线性神经网络的统计保证,这些保证在对数因子范围内与全局极值一致,但适用于 stationary 点及其附近点。这些结果支持神经网络从数学角度看不必必须全局优化的常见观点。我们随后将这些统计保证扩展到浅层 ReLU 神经网络,假设 stationary 网络和目标的第一层权重矩阵几乎相同。更一般而言,尽管目前仅限于浅层神经网络,我们的理论在用数学语言描述神经网络的实用性质方面迈出了重要一步。

英文摘要:

Since statistical guarantees for neural networks are usually restricted to global optima of intricate objective functions, it is unclear whether these theories explain the performances of actual outputs of neural network pipelines. The goal of this paper is, therefore, to bring statistical theory closer to practice. We develop statistical guarantees for shallow linear neural networks that coincide up to logarithmic factors with the global optima but apply to stationary points and the points nearby. These results support the common notion that neural networks do not necessarily need to be optimized globally from a mathematical perspective. We then extend our statistical guarantees to shallow ReLU neural networks, assuming the first layer weight matrices are nearly identical for the stationary network and the target. More generally, despite being limited to shallow neural networks for now, our theories make an important step forward in describing the practical properties of neural networks in mathematical terms.

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。

↑