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

Self-organizing maps with asymmetric neighborhood function.

Takaaki Aoki1, Toshio Aoyagi

  • 1Graduate School of Informatics, Kyoto University, Kyoto, Japan. aoki@acs.i.kyoto-u.ac.jp

Neural Computation
|July 26, 2007
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

Feasibility study of an insole-type active assist device for ankle alignment correction during stepping in patients with knee osteoarthritis.

BMC biomedical engineering·2026
Same author

Characteristics of shoulder range of motion, muscle tightness, and streamlined body position in Japanese female para swimmers with unilateral forearm deficiency and in swimmers with vision impairment: a comparative cross-sectional study.

Journal of physical therapy science·2025
Same author

Setting of the Poincaré section for accurately calculating the phase of rhythmic spatiotemporal dynamics.

Physical review. E·2025
Same author

Bayesian estimation of trunk-leg coordination during walking using phase oscillator models.

Neuroscience research·2024
Same author

Interlimb coordination is not strictly controlled during walking.

Communications biology·2024
Same author

Shoe-Type Walking Assist Device with Controllable Subtalar Joint Alignment at Heel Contact.

IEEE ... International Conference on Rehabilitation Robotics : [proceedings]·2023

This study introduces an asymmetric neighborhood function for the self-organizing map (SOM) algorithm to accelerate training on large datasets. The improved method reduces ordering time complexity and minimizes map distortion, enhancing SOM applicability.

Area of Science:

  • Machine Learning
  • Data Mining
  • Artificial Intelligence

Background:

  • Self-organizing maps (SOMs) are unsupervised learning algorithms used for dimensionality reduction and data visualization.
  • Training SOMs on large datasets is computationally intensive due to topological defects that slow down the ordering process.
  • Existing SOM algorithms struggle with efficiency when dealing with extensive data volumes.

Purpose of the Study:

  • To accelerate the training of self-organizing maps (SOMs) for large-scale data analysis.
  • To address the issue of topological defects that impede the ordering process in SOM training.
  • To develop an improved SOM algorithm that maintains map accuracy while enhancing speed.

Main Methods:

  • Implementation of an asymmetric neighborhood function within the SOM algorithm.

Related Experiment Videos

  • Comparison of the asymmetric approach against traditional symmetric neighborhood functions.
  • Development of an improved asymmetric neighborhood function to mitigate map distortion.

    • Main Results:

    • The asymmetric neighborhood function significantly accelerates the SOM ordering process.
    • The training time complexity for a one-dimensional SOM is reduced from O(N^3) to O(N^2).
    • The improved algorithm effectively suppresses map distortion caused by the asymmetric function.

    Conclusions:

    • An asymmetric neighborhood function is an effective strategy for speeding up SOM training.
    • The proposed enhanced SOM algorithm achieves faster ordering without compromising map quality.
    • This advancement broadens the applicability of SOMs to larger and more complex datasets.