Training neural ODEs using fully discretized simultaneous optimization
File(s)
Published version
Author(s)
Shapovalova, Mariia
Tsay, Calvin
Type
Conference Paper
Abstract
Neural Ordinary Differential Equations (Neural ODEs) represent continuous-time dynamics with neural networks, offering advancements for modeling and control tasks. However, training Neural ODEs requires solving differential equations at each epoch, leading to high computational costs. This work investigates simultaneous optimization methods as a faster training alternative. In particular, we employ a collocation-based, fully discretized formulation and use IPOPT—a solver for large-scale nonlinear optimization—to simultaneously optimize collocation coefficients and neural network parameters. Using the Van der Pol Oscillator as a case study, we demonstrate faster convergence compared to traditional training methods. Furthermore, we introduce a decomposition framework utilizing Alternating Direction Method of Multipliers (ADMM) to effectively coordinate sub-models among data batches. Our results show significant potential for (collocation-based) simultaneous Neural ODE training pipelines.
Date Issued
2025-08-13
Date Acceptance
2025-06-01
Citation
IFAC-PapersOnLine, 2025, 59 (6), pp.469-474
ISSN
2405-8963
Publisher
Elsevier BV
Start Page
469
End Page
474
Journal / Book Title
IFAC-PapersOnLine
Volume
59
Issue
6
Copyright Statement
Copyright © 2025 The Authors. This is an open access article under the CC BY-NC-ND license. Peer review under responsibility of International Federation of Automatic Control.
Identifier
10.1016/j.ifacol.2025.07.190
Source
14th IFAC Symposium on Dynamics and Control of Process Systems, including Biosystems DYCOPS 2025
Subjects
Simultaneous dynamic optimization
nonlinear system identification
neural ODEs
Publication Status
Published
Start Date
2025-06-16
Finish Date
2025-06-19
Coverage Spatial
Bratislava, Slovakia
Date Publish Online
2025-08-13