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

Reinforcement learning with via-point representation.

Hiroyuki Miyamoto1, Jun Morimoto, Kenji Doya

  • 1Kawato Dynamic Brain Project, Japan Science and Technology Corporation, Kyoto, Japan. miyamo@brain.kyutech.ac.jp

Neural Networks : the Official Journal of the International Neural Network Society
|March 24, 2004
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

Minimally invasive pulmonary valve replacement via left anterior mini-thoracotomy using an automated fastener-assisted suturing technique: a case report.

Journal of surgical case reports·2026
Same author

Myoclonin1 haploinsufficiency in motile ciliated cells partially recapitulates epileptic features of Efhc1-deficient mice in adult age.

Molecular and cellular neurosciences·2026
Same author

Adversarial AI reveals mechanisms and treatments for disorders of consciousness.

Nature neuroscience·2026
Same author

Towards natural stand-up movement support: guiding higher-dimensional muscle activation using a Lower-DOF assistive chair.

Frontiers in bioengineering and biotechnology·2026
Same author

Rapid functional reorganization of the targeted contralesional hemisphere induced by one week of noninvasive closed-loop neurofeedback guides motor recovery in post-stroke patients with chronic motor impairment: a phase I trial.

Communications medicine·2026
Same author

Consensus Paper: Models of Cerebellar Functions.

Cerebellum (London, England)·2026
Same journal

Q-learning based asynchronous Boolean control networks stabilization with data loss.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

New results on prescribed-time synchronization of complex networks via intermittent control.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Variance-constrained multi-view ensemble broad network for imbalanced data.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Dynamic analysis and reliable mechanical optimization application of ring HNN effected with a memristive neuron.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

DAFF-Net: A detection and search method for small-scale low surface brightness galaxies.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Quasi-synchronization for complex networks with hybrid pinning intermittent control.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

This study introduces a novel hierarchical reinforcement learning framework for motor control, integrating via-point representation for macro-actions and trajectory generation for primitive actions. The new model successfully controlled simulated robotic systems.

Area of Science:

  • Robotics
  • Machine Learning
  • Control Theory

Background:

  • Conventional reinforcement learning (RL) has been applied to motor control tasks like robot locomotion.
  • Hierarchical architectures are increasingly explored for managing complex control sequences across multiple scales.
  • Existing RL methods often lack the flexibility to modify ongoing movements efficiently.

Purpose of the Study:

  • To propose a new two-level hierarchical learning framework for motor control.
  • To integrate via-point representation and trajectory generation for enhanced control capabilities.
  • To demonstrate the framework's effectiveness in modifying ongoing movements.

Main Methods:

  • Developed a hierarchical framework with two levels: a higher level using via-point representation for macro-actions and a lower level using a trajectory generator for primitive actions.

Related Experiment Videos

  • Implemented the framework using reinforcement learning principles.
  • Utilized computer simulation for testing, specifically the cart-pole swing-up task.

    • Main Results:

    • The proposed framework successfully executed the cart-pole swing-up task in computer simulations.
    • The system demonstrated the ability to modify ongoing movements through temporally localized via-points and trajectory generation.
    • Achieved successful control sequences within the simulated environment.

    Conclusions:

    • The novel hierarchical learning framework effectively addresses motor control challenges.
    • Via-point representation and trajectory generation provide a robust approach for complex robotic movements.
    • The framework shows promise for advanced applications in robotics and autonomous systems.