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 Experiment Videos

Learning in recurrent finite difference networks

F S Tsung1, G W Cottrell

  • 1Chung Tai Ch'an Temple, Taiwan, Republic of China.

International Journal of Neural Systems
|September 1, 1995
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same author

Ultrabright Electron Bunch Injection in a Plasma Wakefield Driven by a Superluminal Flying Focus Electron Beam.

Physical review letters·2022
Same author

Electron acceleration at oblique angles via stimulated Raman scattering at laser irradiance >10^{16}Wcm^{-2}μm^{2}.

Physical review. E·2021
Same author

Three-dimensional particle-in-cell modeling of parametric instabilities near the quarter-critical density in plasmas.

Physical review. E·2019
Same author

Role of Direct Laser Acceleration of Electrons in a Laser Wakefield Accelerator with Ionization Injection.

Physical review letters·2017
Same author

Formation of Ultrarelativistic Electron Rings from a Laser-Wakefield Accelerator.

Physical review letters·2015
Same author

Improving the self-guiding of an ultraintense laser by tailoring its longitudinal profile.

Physical review letters·2014
Same journal

Latent Space Projections and Atlases, a Cautionary Tale in Deep Neuroimaging using Autoencoders.

International journal of neural systems·2026
Same journal

Transformer-Based Anomaly Detection for Neurodegenerative Screening in MRI Images.

International journal of neural systems·2026
Same journal

Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction.

International journal of neural systems·2026
Same journal

Automatic Seizure Detection using Hierarchical Spectral-Temporal Feature Learning with an Imbalance-Aware Transformer.

International journal of neural systems·2026
Same journal

Pyramid Vision Transformer-Enhanced Conformer Network for Epileptic Seizure Recognition Using MultiChannel EEG Signals.

International journal of neural systems·2026
Same journal

A Time-Frequency Decoupled Contrastive Learning Framework for Electroencephalography-Based Parkinson's Disease Diagnosis.

International journal of neural systems·2026
See all related articles

A new recurrent learning algorithm uses finite difference methods for neural networks. This approach simplifies discrete algorithms while enabling networks to learn complex oscillations, overcoming previous limitations.

Area of Science:

  • Computational Neuroscience
  • Machine Learning

Background:

  • Discrete neural network algorithms often struggle with learning smooth oscillations, especially those with large periods.
  • Existing methods can lead to waveform distortion or complete learning failure in such cases.

Purpose of the Study:

  • To derive a recurrent learning algorithm based on finite difference discretization of continuous neural network equations.
  • To address the limitations of discrete algorithms in learning complex oscillatory behaviors.

Main Methods:

  • Developed a novel algorithm by applying finite difference discretization to continuous neural network equations.
  • Analyzed the algorithm's ability to learn smooth oscillations with large periods.
  • Derived formulas for learning time constants and time delays within this framework.

Related Experiment Videos

Main Results:

  • The finite difference formulation simplifies discrete algorithms while preserving essential characteristics of continuous equations.
  • Successfully explained and provided a method to overcome the difficulty in learning large-period oscillations.
  • Presented formulas for learning time constants and time delays.

Conclusions:

  • The proposed finite difference-based recurrent learning algorithm offers a robust solution for training neural networks on complex oscillatory dynamics.
  • This method enhances the learning capabilities of discrete neural networks, particularly for tasks involving smooth, large-period oscillations.