Modeling Time Series Data with Ordinary Differential Equations and Neural Networks

RESOURCE

Abstract

This suite of code learns neural network models to emulate the dynamics of time series data. To do so, it learns a set of "neural shape functions" that provide a common vocabulary for expressing the various dynamics operators. The code is implemented in Python and TensorFlow.
Developers:
Desautels, Thomas [1] Reeves, Majerle [2]
  1. Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
  2. UC Merced
Release Date:
2019-08-09
Project Type:
Open Source, Publicly Available Repository
Software Type:
Scientific
Programming Languages:
Python
Version:
0.0.1
Licenses:
MIT License
Sponsoring Org.:
Code ID:
31120
Site Accession Number:
989386
Research Org.:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
Country of Origin:
United States

RESOURCE

Citation Formats

Desautels, Thomas, and Reeves, Majerle. Modeling Time Series Data with Ordinary Differential Equations and Neural Networks. Computer Software. https://github.com/tadesautels/PSINN. USDOE National Nuclear Security Administration (NNSA). 09 Aug. 2019. Web. doi:10.11578/dc.20191011.2.
Desautels, Thomas, & Reeves, Majerle. (2019, August 09). Modeling Time Series Data with Ordinary Differential Equations and Neural Networks. [Computer software]. https://github.com/tadesautels/PSINN. https://doi.org/10.11578/dc.20191011.2.
Desautels, Thomas, and Reeves, Majerle. "Modeling Time Series Data with Ordinary Differential Equations and Neural Networks." Computer software. August 09, 2019. https://github.com/tadesautels/PSINN. https://doi.org/10.11578/dc.20191011.2.
@misc{ doecode_31120,
title = {Modeling Time Series Data with Ordinary Differential Equations and Neural Networks},
author = {Desautels, Thomas and Reeves, Majerle},
abstractNote = {This suite of code learns neural network models to emulate the dynamics of time series data. To do so, it learns a set of "neural shape functions" that provide a common vocabulary for expressing the various dynamics operators. The code is implemented in Python and TensorFlow.},
doi = {10.11578/dc.20191011.2},
url = {https://doi.org/10.11578/dc.20191011.2},
howpublished = {[Computer Software] \url{https://doi.org/10.11578/dc.20191011.2}},
year = {2019},
month = {aug}
}