arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1902.01028cs.LGcs.DScs.NEmath.OCstat.ML
收藏

SGD能学习具有可证明泛化的循环神经网络吗?

Can SGD Learn Recurrent Neural Networks with Provable Generalization?

Microsoft Research AI(微软研究院AI);Stanford University(斯坦福大学)

Zeyuan Allen-Zhu, Yuanzhi Li

摘要:

循环神经网络(RNN)是序列数据分析中最流行的模型之一。然而,在基础PAC学习语言中,它能学习什么概念类?此外,同一个递归单元如何同时从不同的输入令牌学习函数到不同的输出令牌,而不会相互影响?RNN的现有泛化范围随输入长度呈指数级增长,大大限制了其实际意义。在本文中,我们使用香草随机梯度下降(SGD),RNN实际上可以有效地学习一些值得注意的概念类,这意味着时间和样本复杂度在输入长度上都是多项式(或几乎是多项式,取决于概念)。该概念类至少包括这样的函数,其中每个输出令牌使用平滑的两层神经网络从先前令牌的输入生成。

英文摘要:

Recurrent Neural Networks (RNNs) are among the most popular models in sequential data analysis. Yet, in the foundational PAC learning language, what concept class can it learn? Moreover, how can the same recurrent unit simultaneously learn functions from different input tokens to different output tokens, without affecting each other? Existing generalization bounds for RNN scale exponentially with the input length, significantly limiting their practical implications. In this paper, we show using the vanilla stochastic gradient descent (SGD), RNN can actually learn some notable concept class efficiently, meaning that both time and sample complexity scale polynomially in the input length (or almost polynomially, depending on the concept). This concept class at least includes functions where each output token is generated from inputs of earlier tokens using a smooth two-layer neural network.

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。

↑