Generalization bounds for deep convolutional neural networks
Machine Learning
2020-04-09 v6 Artificial Intelligence
Neural and Evolutionary Computing
Statistics Theory
Machine Learning
Statistics Theory
Abstract
We prove bounds on the generalization error of convolutional networks. The bounds are in terms of the training loss, the number of parameters, the Lipschitz constant of the loss and the distance from the weights to the initial weights. They are independent of the number of pixels in the input, and the height and width of hidden feature maps. We present experiments using CIFAR-10 with varying hyperparameters of a deep convolutional network, comparing our bounds with practical generalization gaps.
Cite
@article{arxiv.1905.12600,
title = {Generalization bounds for deep convolutional neural networks},
author = {Philip M. Long and Hanie Sedghi},
journal= {arXiv preprint arXiv:1905.12600},
year = {2020}
}
Comments
Published as a conference paper at ICLR 2020