We utilize a connection between compositional kernels and branching processes via Mehler’s formula to study deep neural networks. This new probabilistic insight provides us a novel perspective on the mathematical role of activation functions in compositional neural networks. We study the unscaled and rescaled limits of the compositional kernels and explore the different phases of the limiting behavior, as the compositional depth increases. We investigate the memorization capacity of the compositional kernels and neural networks by characterizing the interplay among compositional depth, sample size, dimensionality, and non-linearity of the activation. Explicit formulas on the eigenvalues of the compositional kernel are provided, which quantify the complexity of the corresponding reproducing kernel Hilbert space. On the methodological front, we propose a new random features algorithm, which compresses the compositional layers by devising a new activation function.

More Research From These Scholars

BFI Working Paper Oct 16, 2020

A Precise High-Dimensional Asymptotic Theory for Boosting and Minimum-L1-Norm Interpolated Classifiers

Tengyuan Liang, Pragya Sur
Topics:  Technology & Innovation
BFI Working Paper Oct 16, 2020

Estimating Certain Integral Probability Metrics (IPMs) Is as Hard as Estimating under the IPMs

Tengyuan Liang
Topics:  Technology & Innovation
BFI Working Paper Oct 16, 2020

How Well Generative Adversarial Networks Learn Distributions

Tengyuan Liang
Topics:  Technology & Innovation