A Training-Free Conditional Diffusion Model for Learning Stochastic Dynamical Systems
SIAM Journal on Scientific Computing · 31.10.2025 · 47 (5) · C1144-C1171
DOI: 10.1137/24m1699589 ↗Kurz gefasst
Fachartikel in der Zeitschrift SIAM Journal on Scientific Computing, online erschienen am 13.10.2025. Verfasst von Yanfang Liu und 3 weiteren Autor:innen. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 42 Literaturangaben. Gefördert von: Office of Science, Office of Advanced Scientific Computing Research, UT-Battelle, LLC, Air Force Office of Scientific Research.
Abstract (Originalsprache)
This study introduces a training-free conditional diffusion model for learning unknown SDEs using data. The proposed approach addresses key challenges in computational efficiency and accuracy for modeling SDEs by utilizing a score-based diffusion model to approximate their stochastic flow map. Unlike the existing methods, this technique is based on an analytically derived closed-form exact score function, which can be efficiently estimated by Monte Carlo methods using the trajectory data, and eliminates the need for neural network training to learn the score function. By generating labeled data through solving the corresponding reverse ODE, the approach enables supervised learning of the flow map. Extensive numerical experiments across various SDE types, including linear, nonlinear, and multidimensional systems, demonstrate the versatility and effectiveness of the method. The learned models exhibit significant improvements in predicting both short-term and long-term behaviors of unknown stochastic systems, often surpassing baseline methods such as generative adversarial networks in estimating drift and diffusion coefficients. Reproducibility of computational results. This paper has been awarded the “SIAM Reproducibility Badge: Code and data available” as a recognition that the authors have followed reproducibility principles valued by SISC and the scientific computing community. Code and data that allow readers to reproduce the results in this paper are available at https://github.com/YanfangLiu11/Conditional-Diffusion-Model-for-SDE-Learning and in the supplementary materials ( Conditional-Diffusion-Model-for-SDE-Learning-main.zip [235KB]). [Formula: see text]
Förderung
- Office of Science, Office of Advanced Scientific Computing Research
- UT-Battelle, LLC
- Air Force Office of Scientific Research
Zitieren
APA
Liu, Y., Chen, Y., Xiu, D., & Zhang, G. (2025). A Training-Free Conditional Diffusion Model for Learning Stochastic Dynamical Systems. SIAM Journal on Scientific Computing, 47(5), C1144-C1171. https://doi.org/10.1137/24m1699589
Vancouver
Liu Y, Chen Y, Xiu D, Zhang G. A Training-Free Conditional Diffusion Model for Learning Stochastic Dynamical Systems. SIAM Journal on Scientific Computing. 2025;47(5):C1144-C1171. doi:10.1137/24m1699589
BibTeX
@article{Liu2025,
author = {Liu, Yanfang and Chen, Yuan and Xiu, Dongbin and Zhang, Guannan},
title = {A Training-Free Conditional Diffusion Model for Learning Stochastic Dynamical Systems},
journal = {SIAM Journal on Scientific Computing},
year = {2025},
volume = {47},
number = {5},
pages = {C1144--C1171},
doi = {10.1137/24m1699589}
}
EUVIMED