DOE CODE   /    Search Results   /    pnnl/neural_ODE_ICLR2020

pnnl/neural_ODE_ICLR2020

RESOURCE

Publicly Accessible Repository
https://github.com/pnnl/neural_ODE_ICLR2020
https://doi.org/10.11578/dc.20240614.197

SAVE / SHARE



PNNL Badge

Abstract

We show how to model discrete ordinary differential equations (ODE) with algebraic nonlinearities as deep neural networks with varying degrees of prior knowledge. We derive the stability guarantees of the network layers based on the implicit constraints imposed on the weight's eigenvalues. Moreover, we show how to use barrier methods to generically handle additional inequality constraints. We demonstrate the prediction accuracy of learned neural ODEs evaluated on open-loop simulations compared to ground truth dynamics with bi-linear terms.
Developers:
Tuor, Aaron [1] ;  Drgona, Jan [2]
  1. Pacific Northwest National Laboratory
  2. Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
Release Date:
2021-11-16
Project Type:
Open Source, Publicly Available Repository
Software Type:
Scientific
Version:
1
Licenses:
BSD 3-clause "New" or "Revised" License
Sponsoring Org.:
Code ID:
67085
Site Accession Number:
Battelle IPID 31922-E
Research Org.:
Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
Country of Origin:
United States

RESOURCE

Publicly Accessible Repository
https://github.com/pnnl/neural_ODE_ICLR2020
https://doi.org/10.11578/dc.20240614.197

SAVE / SHARE



PNNL Badge

Citation Formats

Tuor, Aaron, and Drgona, Jan. pnnl/neural_ODE_ICLR2020. Computer Software. https://github.com/pnnl/neural_ODE_ICLR2020. USDOE. 16 Nov. 2021. Web. doi:10.11578/dc.20240614.197.
Tuor, Aaron, & Drgona, Jan. (2021, November 16). pnnl/neural_ODE_ICLR2020. [Computer software]. https://github.com/pnnl/neural_ODE_ICLR2020. https://doi.org/10.11578/dc.20240614.197.
Tuor, Aaron, and Drgona, Jan. "pnnl/neural_ODE_ICLR2020." Computer software. November 16, 2021. https://github.com/pnnl/neural_ODE_ICLR2020. https://doi.org/10.11578/dc.20240614.197.
@misc{ doecode_67085,
title = {pnnl/neural_ODE_ICLR2020},
author = {Tuor, Aaron and Drgona, Jan},
abstractNote = {We show how to model discrete ordinary differential equations (ODE) with algebraic nonlinearities as deep neural networks with varying degrees of prior knowledge. We derive the stability guarantees of the network layers based on the implicit constraints imposed on the weight's eigenvalues. Moreover, we show how to use barrier methods to generically handle additional inequality constraints. We demonstrate the prediction accuracy of learned neural ODEs evaluated on open-loop simulations compared to ground truth dynamics with bi-linear terms.},
doi = {10.11578/dc.20240614.197},
url = {https://doi.org/10.11578/dc.20240614.197},
howpublished = {[Computer Software] \url{https://doi.org/10.11578/dc.20240614.197}},
year = {2021},
month = {nov}
}