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

Neural network control of simple limb movements.

H C Kwan1, T H Yeap, B C Jiang

  • 1Department of Physiology, University of Toronto, Ont., Canada.

Canadian Journal of Physiology and Pharmacology
|January 1, 1990
PubMed
Summary

A small neural network can learn and control single-joint movements. Adjusting synaptic weights allows generalization to new movement patterns, suggesting network relaxation is key to movement control.

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

Crepidiastrum denticulatum Extract Ameliorates Kidney Ischemia-Reperfusion Injury in Mice.

Transplantation proceedings·2018
Same author

Protective Effect of Eupatilin Pretreatment Against Hepatic Ischemia-Reperfusion Injury in Mice.

Transplantation proceedings·2016
Same author

5,7-dihydroxy-3,4,6-trimethoxyflavone attenuates ischemic damage and apoptosis in mouse islets.

Transplantation proceedings·2015
Same author

Aerobic and resistance exercise training program intervention for enhancing gait function in elderly and chronically ill Taiwanese patients.

Public health·2015
Same author

Protective effect of eupatilin against renal ischemia-reperfusion injury in mice.

Transplantation proceedings·2015
Same author

Global comparison of gene expression between subcutaneous and intramuscular adipose tissue of mature Erhualian pig.

Genetics and molecular research : GMR·2013

Area of Science:

  • Computational neuroscience
  • Robotics
  • Biophysics

Background:

  • Understanding the neural basis of motor control is crucial for developing advanced prosthetics and robots.
  • Previous models often lack the ability to generalize learned movements to new, unlearned trajectories.

Purpose of the Study:

  • To investigate the computational principles underlying motor control using a simplified neural network model.
  • To demonstrate how a non-linear neural network can learn and generalize single-joint movement control.

Main Methods:

  • A small non-linear network of neuron-like elements was designed.
  • Synaptic weights were adjusted through a learning process to embed control of movement speeds.
  • The network's ability to generalize learned trajectories was tested.

Related Experiment Videos

Main Results:

  • The network successfully embedded control for single-joint movements at various speeds.
  • Learning of specific movement trajectories generalized to a family of unlearned trajectories.
  • Network relaxation dynamics were observed to correlate with movement generation.

Conclusions:

  • A small non-linear neural network can learn and generalize motor control.
  • Synaptic weight adjustment is a key mechanism for learning movement patterns.
  • Network relaxation dynamics are hypothesized to be both causal and computational for movement generation and control.