Jove
Visualize
Contact Us

Related Experiment Videos

Continuous-time temporal back-propagation with adaptable time delays.

S P Day1, M R Davenport

  • 1British Columbia Univ., Vancouver, BC.

IEEE Transactions on Neural Networks
|January 1, 1993
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

A segmented ring, cylindrical ion trap source for time-of-flight mass spectrometry.

Journal of the American Society for Mass Spectrometry·2013
Same author

Remodelling of cerebrospinal fluid lipoproteins after subarachnoid hemorrhage.

Atherosclerosis·2003
Same author

Platelet aggregatory responses to low-dose collagen are maintained in hirudin-anticoagulated whole blood for 24 h when stored at room temperature.

Platelets·2001
Same author

Quantitative detection of platelet aggregates in whole blood without fixation.

Platelets·2000
Same author

Heparin-induced platelet dysfunction and cardiopulmonary bypass.

The Annals of thoracic surgery·2000
Same author

Role of genotypic analysis of the thymidine kinase gene of herpes simplex virus for determination of neurovirulence and resistance to acyclovir.

Journal of clinical microbiology·1999
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles
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

This study introduces an extended backpropagation technique for continuous-time feedforward networks with adaptable time delays, improving signal prediction accuracy for chaotic systems like the Mackey-Glass signal.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Neural Networks

Background:

  • Conventional backpropagation networks struggle with true signal prediction and adaptability in dynamic systems.
  • Time delays in neural networks are crucial for processing temporal information but are often fixed.

Purpose of the Study:

  • To extend backpropagation for continuous-time feedforward networks incorporating adaptable internal time delays.
  • To enhance the capability of neural networks for accurate signal prediction and spatiotemporal pattern recognition.

Main Methods:

  • Developed a novel backpropagation algorithm for feedforward networks with adaptable, continuous time delays.
  • Trained networks using continuous multidimensional signals, suitable for parallel hardware implementation.
  • Simulated network performance on predicting the Mackey-Glass chaotic signal.

Related Experiment Videos

Main Results:

  • Networks with adaptable delays achieved significantly lower prediction errors compared to fixed-delay and conventional networks.
  • Adaptable delay networks reduced prediction error by over 50% compared to fixed-delay networks.
  • The trained networks demonstrated the ability to autonomously generate approximations of the input chaotic signal.

Conclusions:

  • The extended backpropagation technique effectively handles adaptable time delays in continuous-time networks.
  • This method offers superior performance in true signal prediction tasks, particularly for chaotic time series.
  • The developed networks are suitable for applications in signal prediction, production, and spatiotemporal pattern recognition.