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 2002.08335stat.MLcs.LG
收藏

核流神经网络内层的深度正则化和直接训练

Deep regularization and direct training of the inner layers of Neural Networks with Kernel Flows

Caltech(加州理工学院)

Gene Ryan Yoo, Houman Owhadi

摘要:

提出了一种基于核流的人工神经网络正则化方法。KFs被引入作为回归/克里金法中的核选择方法,其基础是将随机批次数据集中的插值点数量减半所导致的准确性损失最小化。编写用于ANN的组成结构的函数表示的$f_θ(x) = \big(f^{(n)}_{θ_n}\circ f^{(n-1)}_{θ_{n-1}} \circ \dots \circ f^{(1)}_{θ_1}\big)(x)$,内层输出$h^{(i)}(x) = \big(f^{(i)}_{θ_i}\circ f^{(i-1)}_{θ_{i-1}} \circ \dots \circ f^{(1)}_{θ_1}\big)(x)$定义特征图和内核$k^{(i)}(x,x')=\exp(- γ_i \|h^{(i)}(x)-h^{(i)}(x')\|_2^2)$的层次结构。当与一批数据集组合时,这些内核产生KF损失$e_2^{(i)}$(通过使用随机的一半批次来预测另一半批次而产生的$L^2$回归误差),这取决于内层$θ_1,\ldots,θ_i$(和$γ_i$)的参数。所提出的方法简单地包括在聚合这些KF损失的一个子集与一个经典的输出损失。我们在CNN和WRN上测试了所提出的方法,而没有改变结构或输出分类器,并报告减少了测试错误,减少了泛化差距,并增加了对分布偏移的鲁棒性,而不会显着增加计算复杂性。我们怀疑这些结果可能是由这样一个事实解释的,即虽然传统的训练只采用线性泛函,(广义矩)的经验分布定义的数据集,并可能容易陷入神经切线核制度(在过度参数化下),提出的损失函数(定义为经验分布的非线性泛函)有效地训练由CNN定义的底层内核,而不是用该内核回归数据。

英文摘要:

We introduce a new regularization method for Artificial Neural Networks (ANNs) based on Kernel Flows (KFs). KFs were introduced as a method for kernel selection in regression/kriging based on the minimization of the loss of accuracy incurred by halving the number of interpolation points in random batches of the dataset. Writing $f_θ(x) = \big(f^{(n)}_{θ_n}\circ f^{(n-1)}_{θ_{n-1}} \circ \dots \circ f^{(1)}_{θ_1}\big)(x)$ for the functional representation of compositional structure of the ANN, the inner layers outputs $h^{(i)}(x) = \big(f^{(i)}_{θ_i}\circ f^{(i-1)}_{θ_{i-1}} \circ \dots \circ f^{(1)}_{θ_1}\big)(x)$ define a hierarchy of feature maps and kernels $k^{(i)}(x,x')=\exp(- γ_i \|h^{(i)}(x)-h^{(i)}(x')\|_2^2)$. When combined with a batch of the dataset these kernels produce KF losses $e_2^{(i)}$ (the $L^2$ regression error incurred by using a random half of the batch to predict the other half) depending on parameters of inner layers $θ_1,\ldots,θ_i$ (and $γ_i$). The proposed method simply consists in aggregating a subset of these KF losses with a classical output loss. We test the proposed method on CNNs and WRNs without alteration of structure nor output classifier and report reduced test errors, decreased generalization gaps, and increased robustness to distribution shift without significant increase in computational complexity. We suspect that these results might be explained by the fact that while conventional training only employs a linear functional (a generalized moment) of the empirical distribution defined by the dataset and can be prone to trapping in the Neural Tangent Kernel regime (under over-parameterizations), the proposed loss function (defined as a nonlinear functional of the empirical distribution) effectively trains the underlying kernel defined by the CNN beyond regressing the data with that kernel.

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。

↑