Open AccessClassification with Runge-Kutta networks and feature space augmentation
Author(s) -
Elisa Giesecke,
Axel Kröner
Publication year - 2021
Publication title -
journal of computational dynamics
Language(s) - English
Resource type - Journals
eISSN - 2158-2505
pISSN - 2158-2491
DOI - 10.3934/jcd.2021018
Subject(s) - runge–kutta methods , artificial neural network , space (punctuation) , point (geometry) , feature (linguistics) , computer science , artificial intelligence , image (mathematics) , mathematics , algorithm , geometry , numerical analysis , mathematical analysis , linguistics , philosophy , operating system
In this paper we combine an approach based on Runge-Kutta Nets considered in [ Benning et al., J. Comput. Dynamics, 9, 2019 ] and a technique on augmenting the input space in [ Dupont et al., NeurIPS , 2019] to obtain network architectures which show a better numerical performance for deep neural networks in point and image classification problems. The approach is illustrated with several examples implemented in PyTorch.