Title: Deep Networks with Adaptive Nyström Approximation Show Chinese title

Title: 深度网络与自适应 Nyström 近似

Abstract: Recent work has focused on combining kernel methods and deep learning to exploit the best of the two approaches. Here, we introduce a new architecture of neural networks in which we replace the top dense layers of standard convolutional architectures with an approximation of a kernel function by relying on the Nystr{\"o}m approximation. Our approach is easy and highly flexible. It is compatible with any kernel function and it allows exploiting multiple kernels. We show that our architecture has the same performance than standard architecture on datasets like SVHN and CIFAR100. One benefit of the method lies in its limited number of learnable parameters which makes it particularly suited for small training set sizes, e.g. from 5 to 20 samples per class. Show Chinese abstract

Abstract: 近期的工作集中在结合核方法和深度学习以利用两种方法的优点。 在这里，我们引入了一种新的神经网络架构，在该架构中，我们用基于 Nyström 近似的核函数来替代标准卷积架构的顶层密集层。 我们的方法简单且高度灵活。 它兼容任何核函数，并允许使用多个核。 我们证明了我们的架构在像 SVHN 和 CIFAR100 这样的数据集上与标准架构具有相同的性能。 该方法的一个好处在于其有限的学习参数数量，这使得它特别适合于小训练集大小的情况，例如每类样本从 5 到 20 个样本。