Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Model Approaches for Pharmacokinetic Data: Distributed Parameter Models01:06

Model Approaches for Pharmacokinetic Data: Distributed Parameter Models

148
Pharmacokinetic models are mathematical constructs that represent and predict the time course of drug concentrations in the body, providing meaningful pharmacokinetic parameters. These models are categorized into compartment, physiological, and distributed parameter models.
The distributed parameter models are specifically designed to account for variations and differences in some drug classes. This model is particularly useful for assessing regional concentrations of anticancer or...
148
Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving01:29

Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving

126
Mechanistic models play a crucial role in algorithms for numerical problem-solving, particularly in nonlinear mixed effects modeling (NMEM). These models aim to minimize specific objective functions by evaluating various parameter estimates, leading to the development of systematic algorithms. In some cases, linearization techniques approximate the model using linear equations.
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...
126

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Survivin, the promising target in hepatocellular carcinoma gene therapy.

Cancer biology & therapy·2008
Same author

Curcumin protects dopaminergic neuron against LPS induced neurotoxicity in primary rat neuron/glia culture.

Neurochemical research·2008
Same author

Cellular mechanisms of reduced sarcoplasmic reticulum Ca2+ content in L-thyroxin induced rat ventricular hypertrophy.

Acta pharmacologica Sinica·2008
Same author

Promoting the formation and stabilization of G-quadruplex by dinuclear RuII complex Ru2(obip)L4.

Inorganic chemistry·2008
Same author

Identification of direct target genes using joint sequence and expression likelihood with application to DAF-16.

PloS one·2008
Same author

In vitro and in vivo investigations on the antiviral activity of a series of mixed-valence rare earth borotungstate heteropoly blues.

European journal of medicinal chemistry·2008

Related Experiment Video

Updated: Oct 18, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.5K

An improved data-free surrogate model for solving partial differential equations using deep neural networks.

Xinhai Chen1,2, Rongliang Chen3, Qian Wan1

  • 1Science and Technology on Parallel and Distributed Processing Laboratory, National University of Defense Technology, Changsha, 410000, China.

Scientific Reports
|October 1, 2021
PubMed
Summary
This summary is machine-generated.

DFS-Net, a novel data-free surrogate model, enhances deep neural network predictions for partial differential equations (PDEs). This physics-informed neural network (PINN) alternative improves accuracy and efficiency without needing simulation data.

More Related Videos

Surrogate Model Development for Digital Experiments in Welding
09:17

Surrogate Model Development for Digital Experiments in Welding

Published on: March 28, 2025

1.3K

Related Experiment Videos

Last Updated: Oct 18, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.5K
Surrogate Model Development for Digital Experiments in Welding
09:17

Surrogate Model Development for Digital Experiments in Welding

Published on: March 28, 2025

1.3K

Area of Science:

  • Computational Science and Engineering
  • Applied Mathematics
  • Artificial Intelligence

Background:

  • Partial differential equations (PDEs) are fundamental in science and engineering.
  • Traditional numerical methods for PDEs are computationally intensive and require mesh generation.
  • Deep neural networks offer potential for efficient surrogate modeling due to their approximation capabilities.

Purpose of the Study:

  • To develop an improved, data-free surrogate model for solving PDEs.
  • To enhance the stability and accuracy of physics-informed neural networks (PINNs).
  • To introduce an attention-based neural network architecture for direct PDE solution approximation.

Main Methods:

  • Proposed DFS-Net, an attention-based neural network architecture.
  • Input expanded spatial and temporal coordinates to directly output observables.
  • Minimized weighted residuals of governing equations and data-fit terms, requiring no external data.

Main Results:

  • DFS-Net demonstrated a favorable balance between accuracy and computational efficiency.
  • The model outperformed existing surrogate models on benchmark PDEs (Helmholtz, Klein-Gordon, Navier-Stokes).
  • Attention mechanism in DFS-Net mitigated prediction instability and inaccuracy issues common in PINNs.

Conclusions:

  • DFS-Net provides a robust and efficient data-free surrogate modeling approach for PDEs.
  • The attention-based mechanism significantly improves the performance of physics-informed neural networks.
  • DFS-Net offers a promising alternative to traditional numerical methods and standard surrogate models.