Rethinking Layer-wise Feature Amounts in Convolutional Neural Network Architectures

Accepted in NeurIPS CRACT 2018, Canada

Martin Mundt, Sagnik Majumder, Tobias Weis, Visvanathan Ramesh, "Rethinking Layer-wise Feature Amounts in Convolutional Neural Network Architectures" In: International Conference on Neural Information Processing Systems (NeurIPS) 2018, Critiquing and Correcting Trends in Machine Learning (CRACT) Workshop. https://arxiv.org/abs/1812.05836

Abstract: We characterize convolutional neural networks with respect to the relative amount of features per layer. Using a skew normal distribution as a parametrized framework, we investigate the common assumption of monotonously increasing feature-counts with higher layers of architecture designs. Our evaluation on models with VGG- type layers on the MNIST, Fashion-MNIST and CIFAR-10 image classification benchmarks provides evidence that motivates rethinking of our common assump- tion: architectures that favor larger early layers seem to yield better accuracy.

Download paper here