当前位置:
X-MOL 学术
›
Comput. Math. Math. Phys.
›
论文详情
Our official English website, www.x-mol.net, welcomes your
feedback! (Note: you will need to create a separate account there.)
Reduced-Order Modeling of Deep Neural Networks
Computational Mathematics and Mathematical Physics ( IF 0.7 ) Pub Date : 2021-07-01 , DOI: 10.1134/s0965542521050109 J. Gusak , T. Daulbaev , E. Ponomarev , A. Cichocki , I. Oseledets
中文翻译:
深度神经网络的降阶建模
更新日期:2021-07-02
Computational Mathematics and Mathematical Physics ( IF 0.7 ) Pub Date : 2021-07-01 , DOI: 10.1134/s0965542521050109 J. Gusak , T. Daulbaev , E. Ponomarev , A. Cichocki , I. Oseledets
Abstract
We introduce a new method for speeding up the inference of deep neural networks. It is somewhat inspired by the reduced-order modeling techniques for dynamical systems. The cornerstone of the proposed method is the maximum volume algorithm. We demonstrate efficiency on neural networks pre-trained on different datasets. We show that in many practical cases it is possible to replace convolutional layers with much smaller fully-connected layers with a relatively small drop in accuracy.
中文翻译:
深度神经网络的降阶建模
摘要
我们介绍了一种加速深度神经网络推理的新方法。它的灵感来自于动态系统的降阶建模技术。所提出方法的基石是最大体积算法。我们展示了在不同数据集上预训练的神经网络的效率。我们表明,在许多实际情况下,可以用更小的全连接层替换卷积层,但精度下降相对较小。