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

Constructive neural-network learning algorithms for pattern classification.

R Parekh1, J Yang, V Honavar

  • 1Allstate Research and Planning Center, Menlo Park, CA 94025, USA. rpare@allstate.com

IEEE Transactions on Neural Networks
|February 6, 2008
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

Transfection of IFNalpha in human glioblastoma cells and tumorigenicity in association with induction of PKR and OAS gene expression.

Journal of neurosurgery·1996
Same author

In vitro transcriptional analysis of TyrR-mediated activation of the mtr and tyrP+3 promoters of Escherichia coli.

Journal of bacteriology·1996
Same author

Case report: ranitidine-induced bradycardia in a patient with dextrocardia.

The American journal of the medical sciences·1996
Same author

Antioxidant-sensitive regulation of inflammatory-response genes in Kaposi's sarcoma cells.

Journal of acquired immune deficiency syndromes and human retrovirology : official publication of the International Retrovirology Association·1996
Same author

Cytokine expression in the presence or absence of late airway responses after antigen challenge of sensitized rats.

American journal of respiratory cell and molecular biology·1996
Same author

Injuries of the abdominal aorta and inferior vena cava in association with thoracolumbar fractures: a lethal combination.

The Journal of trauma·1996
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

New constructive learning algorithms, MPyramid-real and MTiling-real, enable neural networks to classify complex real-valued data with multiple categories. These algorithms improve pattern classification accuracy and efficiency.

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Pattern Recognition

Background:

  • Traditional neural network training often requires manual design of network architecture.
  • Existing constructive algorithms typically handle binary inputs and outputs, limiting their application.
  • Need for adaptable algorithms capable of handling real-valued data and multi-class problems.

Purpose of the Study:

  • To introduce MPyramid-real and MTiling-real, novel constructive learning algorithms.
  • To extend constructive learning to real-valued inputs and M-ary classifications.
  • To demonstrate the convergence and practical utility of these new algorithms.

Main Methods:

  • Development of MPyramid-real and MTiling-real algorithms based on pyramid and tiling concepts.

Related Experiment Videos

  • Theoretical proof of algorithm convergence for real-to-M-ary mappings.
  • Empirical evaluation on practical pattern classification tasks.
  • Integration of a local pruning mechanism for network optimization.

    • Main Results:

    • MPyramid-real and MTiling-real successfully learn real-to-M-ary mappings.
    • Algorithms demonstrate convergence to accurate classifications.
    • Empirical results confirm applicability to real-world pattern classification problems.
    • Pruning step effectively reduces redundant neurons in MTiling-real networks.

    Conclusions:

    • MPyramid-real and MTiling-real represent significant advancements in constructive neural network learning.
    • These algorithms offer a robust solution for complex pattern classification tasks with real-valued data.
    • Network optimization through pruning enhances efficiency and generalizability.