We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: GD-VAEs: Geometric Dynamic Variational Autoencoders for Learning Nonlinear Dynamics and Dimension Reductions

Abstract: We develop data-driven methods incorporating geometric and topological information to learn parsimonious representations of nonlinear dynamics from observations. We develop approaches for learning nonlinear state space models of the dynamics for general manifold latent spaces using training strategies related to Variational Autoencoders (VAEs). Our methods are referred to as Geometric Dynamic (GD) Variational Autoencoders (GD-VAEs). We learn encoders and decoders for the system states and evolution based on deep neural network architectures that include general Multilayer Perceptrons (MLPs), Convolutional Neural Networks (CNNs), and Transpose CNNs (T-CNNs). Motivated by problems arising in parameterized PDEs and physics, we investigate the performance of our methods on tasks for learning low dimensional representations of the nonlinear Burgers equations, constrained mechanical systems, and spatial fields of reaction-diffusion systems. GD-VAEs provide methods for obtaining representations for use in diverse learning tasks involving dynamics.
Comments: 15 figures. arXiv admin note: text overlap with arXiv:2012.03448
Subjects: Machine Learning (cs.LG); Dynamical Systems (math.DS); Numerical Analysis (math.NA); Data Analysis, Statistics and Probability (physics.data-an); Machine Learning (stat.ML)
Cite as: arXiv:2206.05183 [cs.LG]
  (or arXiv:2206.05183v2 [cs.LG] for this version)

Submission history

From: Paul Atzberger [view email]
[v1] Fri, 10 Jun 2022 15:23:23 GMT (12819kb,D)
[v2] Tue, 20 Dec 2022 00:17:33 GMT (12819kb,D)

Link back to: arXiv, form interface, contact.