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

An adaptive high-order neural tree for pattern recognition.

G L Foresti1, T Dolso

  • 1Department of Mathematics and Computer Science (DIMI), University of Udine, Via delle Scienze, 208-33100 Udine, Italy. foresti@dimi.uniud.it

IEEE Transactions on Systems, Man, and Cybernetics. Part B, Cybernetics : a Publication of the IEEE Systems, Man, and Cybernetics Society
|September 21, 2004
PubMed
Summary

A novel adaptive high-order neural tree (AHNT) model classifies complex patterns efficiently. This adaptive neural tree approach improves generalization and reduces model complexity for large datasets.

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

Exploiting a geometrically sampled grid in the steered response power algorithm for localization improvement.

The Journal of the Acoustical Society of America·2017
Same author

A hierarchical approach to feature extraction and grouping.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2008
Same author

Generalized neural trees for pattern classification.

IEEE transactions on neural networks·2008
Same author

Visual inspection of sea bottom structures by an autonomous underwater vehicle.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2008
Same author

Invariant feature extraction and neural trees for range surface classification.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2008
Same author

How a visual surveillance system hypothesizes how you behave.

Behavior research methods·2006

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Pattern Recognition

Background:

  • Classifying large, multidimensional datasets presents significant computational challenges.
  • Existing tree models often require complex parameter tuning or statistical tests.
  • High-order perceptrons offer flexible data division but can increase complexity.

Purpose of the Study:

  • To introduce a new neural tree model, the adaptive high-order neural tree (AHNT).
  • To enhance the classification of large, multidimensional pattern sets.
  • To develop a model that automatically adapts its complexity.

Main Methods:

  • The AHNT model recursively partitions training data into subsets assigned to child nodes.
  • Each node utilizes a high-order perceptron (HOP) with an automatically tuned order based on data complexity.

Related Experiment Videos

  • First-order nodes use hyperplanes, while HOPs allow arbitrary input space division.

    • Main Results:

    • The AHNT demonstrates superior generalization compared to trees with homogeneous nodes.
    • The proposed model generates smaller, more parsimonious trees.
    • AHNT avoids the need for complex statistical tests and extensive parameter pre-selection.

    Conclusions:

    • The adaptive high-order neural tree (AHNT) offers an effective solution for multidimensional pattern classification.
    • AHNT provides improved generalization and model simplicity.
    • This approach represents a significant advancement in adaptive neural network architectures.