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 2007.08864cs.LGstat.ML
收藏

固定蝶形结构的稀疏线性网络:理论与实践

Sparse Linear Networks with a Fixed Butterfly Structure: Theory and Practice

Technion Israel Institute of Technology(以色列理工学院)

Nir Ailon, Omer Leibovich, Vineet Nair

摘要:

蝴蝶网络由许多层组成,每层都有线性数量的非零权重(预先指定)。快速Johnson-Lindenstrauss变换(FJLT)可以表示为蝴蝶网络,然后投影到坐标的随机子集上。此外,基于FJLT的随机矩阵以高概率逼近任何矩阵对向量的作用。受这些事实的启发,我们建议用基于蝴蝶网络的架构来取代任何神经网络中的密集线性层。所提出的架构显着改善后,在一个标准的密集层所需的权重的平方数几乎是线性的,在所得到的运营商的可表达性几乎没有妥协。在一系列广泛的实验中,包括对NLP和视觉数据的监督预测,我们表明,这不仅产生了与现有知名架构相匹配的结果,有时甚至优于现有的知名架构,而且还提供了更快的部署训练和预测。为了理解带有蝴蝶网络的神经网络所带来的优化问题,我们还研究了编码器-解码器网络的优化前景,其中编码器被蝴蝶网络取代,然后是较小维度的密集线性层。文中提出的理论结果解释了为什么我们提出的方法不会影响训练速度和结果。

英文摘要:

A butterfly network consists of logarithmically many layers, each with a linear number of non-zero weights (pre-specified). The fast Johnson-Lindenstrauss transform (FJLT) can be represented as a butterfly network followed by a projection onto a random subset of the coordinates. Moreover, a random matrix based on FJLT with high probability approximates the action of any matrix on a vector. Motivated by these facts, we propose to replace a dense linear layer in any neural network by an architecture based on the butterfly network. The proposed architecture significantly improves upon the quadratic number of weights required in a standard dense layer to nearly linear with little compromise in expressibility of the resulting operator. In a collection of wide variety of experiments, including supervised prediction on both the NLP and vision data, we show that this not only produces results that match and at times outperform existing well-known architectures, but it also offers faster training and prediction in deployment. To understand the optimization problems posed by neural networks with a butterfly network, we also study the optimization landscape of the encoder-decoder network, where the encoder is replaced by a butterfly network followed by a dense linear layer in smaller dimension. Theoretical result presented in the paper explains why the training speed and outcome are not compromised by our proposed approach.

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。

↑