Analysis of the rate of convergence of an over-parametrized convolutional neural network image classifier learned by gradient descent

arXiv:2405.07619v1 Announce Type: cross
Abstract: Image classification based on over-parametrized convolutional neural networks with a global average-pooling layer is considered. The weights of the network are learned by gradient descent. A bound on the rate of convergence of the difference between the misclassification risk of the newly introduced convolutional neural network estimate and the minimal possible value is derived.

abstract analysis arxiv classification classifier convergence convolutional convolutional neural network convolutional neural networks cs.lg difference global gradient image layer network networks neural network neural networks pooling rate stat.ml type