Learn one size to infer all: Exploiting translational symmetries in delay-dynamical and spatio-temporal systems using scalable neural networks
Preprint
Abstract
We design scalable neural networks adapted to translational symmetries in dynamical systems, capable of inferring untrained high-dimensional dynamics for different system sizes. We train these networks to predict the dynamics of delay-dynamical and spatio-temporal systems for a single size. Then, we drive the networks by their own predictions. We demonstrate that by scaling the size of the trained network, we can predict the complex dynamics for larger or smaller system sizes. Thus, the network learns from a single example and, by exploiting symmetry properties, infers entire bifurcation diagrams.Keywords
All Related Versions
This publication has 39 references indexed in Scilit:
- Biologically informed deep neural network for prostate cancer discoveryNature, 2021
- Next generation reservoir computingNature Communications, 2021
- Teaching recurrent neural networks to infer global temporal structure from local examplesNature Machine Intelligence, 2021
- Reservoir Computing2021
- Integrating machine learning and multiscale modeling—perspectives, challenges, and opportunities in the biological, biomedical, and behavioral sciencesnpj Digital Medicine, 2019
- Oscillatory dynamics of an intravenous glucose tolerance test model with delay intervalChaos: An Interdisciplinary Journal of Nonlinear Science, 2017
- Climate models with delay differential equationsChaos: An Interdisciplinary Journal of Nonlinear Science, 2017
- Reservoir Computing TrendsKI - Künstliche Intelligenz, 2012
- Information processing using a single dynamical node as complex systemNature Communications, 2011
- Ikeda Hopf bifurcation revisitedPhysica D: Nonlinear Phenomena, 2004