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 topological tree for online vector quantization and data clustering.

Pengfei Xu1, Chip-Hong Chang, Andrew Paplinski

  • 1Centre for High Performance Embedded Systems, Nanyang Technological University, 639798 Singapore. xupengfei@pmail.ntu.edu.sg

IEEE Transactions on Systems, Man, and Cybernetics. Part B, Cybernetics : a Publication of the IEEE Systems, Man, and Cybernetics Society
|June 24, 2005
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

The impact of subchronic ozone exposure on serum metabolome and the mechanisms of abnormal bile acid and arachidonic acid metabolisms in the liver.

Ecotoxicology and environmental safety·2023
Same author

Elabela-APJ axis attenuates cerebral ischemia/reperfusion injury by inhibiting neuronal ferroptosis.

Free radical biology & medicine·2023
Same author

Fast response CdS-CdS<sub>x</sub>Te<sub>1-</sub><sub>x</sub>-CdTe core-shell nanobelt photodetector.

Science bulletin·2023
Same author

GmWAK1, Novel Wall-Associated Protein Kinase, Positively Regulates Response of Soybean to <i>Phytophthora sojae</i> Infection.

International journal of molecular sciences·2023
Same author

Optimized coaxial focused electrohydrodynamic jet printing of highly ordered semiconductor sub-microwire arrays for high-performance organic field-effect transistors.

Nanoscale·2023
Same author

Novel inhibition of AKR1C3 and androgen receptor axis by PTUPB synergizes enzalutamide treatment in advanced prostate cancer.

Oncogene·2023
Same journal

Strategic Ability Updating in Concurrent Games by Coalitional Commitment.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2015
Same journal

Meta-Analysis of the First Facial Expression Recognition Challenge.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

Adjustable model-based fusion method for multispectral and panchromatic images.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

Face Feature Weighted Fusion Based on Fuzzy Membership Degree for Video Face Recognition.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

A New Adaptive Fast Cellular Automaton Neighborhood Detection and Rule Identification Algorithm.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

Human-arm-and-hand-dynamic model with variability analyses for a stylus-based haptic interface.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
See all related articles

A new online self-organizing topological tree (SOTT) offers faster learning and superior topology preservation compared to traditional self-organizing maps (SOM). This efficient vector quantization method excels in large datasets and hierarchical mapping.

Area of Science:

  • Machine Learning
  • Data Mining
  • Artificial Intelligence

Background:

  • Self-Organizing Maps (SOM) perform vector quantization (VQ) and topology-preserving mapping.
  • Classical clustering algorithms face limitations in performance and scalability for large datasets.

Purpose of the Study:

  • To introduce an online self-organizing topological tree (SOTT) with enhanced learning efficiency and topology preservation.
  • To develop a hybrid clustering algorithm leveraging SOTT's online learning and multiresolution capabilities.

Main Methods:

  • Proposed a novel online self-organizing topological tree (SOTT) with a new learning rule.
  • Achieved computational complexity of O(log N) for SOTT, significantly faster than basic SOM's O(N).
  • Devised a hybrid clustering algorithm incorporating an online-updatable linkage metric for accelerated hierarchical clustering.

Related Experiment Videos

Main Results:

  • SOTT demonstrates comparable reconstruction performance to full-search SOM but with substantially reduced computation time.
  • SOTT provides hierarchical mapping and progressive transmission/decoding properties, uncommon in other vector quantizers.
  • The SOTT hybrid clustering algorithm exhibits enhanced performance and memory efficiency, independent of dataset size.

Conclusions:

  • The proposed SOTT offers a superior and more efficient alternative to traditional SOMs for vector quantization and topology preservation.
  • The SOTT hybrid clustering algorithm effectively addresses limitations of classical methods, proving suitable for large-scale data analysis.