Jove
Visualize
Contact Us

Related Experiment Videos

Discrimination networks for maximum selection.

Brijnesh J Jain1, Fritz Wysotzki

  • 1Department of Computer Science, Sekr FR 5-8, Technical University Berlin, Franklinstr 28/29, D-10587 Berlin, Germany. bjj@cs.tu-berlin.de

Neural Networks : the Official Journal of the International Neural Network Society
|December 24, 2003
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

Molecule kernels: a descriptor- and alignment-free quantitative structure-activity relationship approach.

Journal of chemical information and modeling·2008
Same journal

A boundary-regularization-enhanced video anomaly detection network based on context-adaptive spatio-temporal conditional diffusion.

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

MT<sup>2</sup>-CSD and LLM-CRAN: A new dataset and an LLM-based multi-semantic knowledge fusion model for conversational stance detection.

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

TriAlignNet: A triple-path cross-modality alignment framework for multimodal time series forecasting.

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

Anchor-based disentanglement framework for incremental multi-view clustering.

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

Complex-valued amplitude-phase interference modeling for adversarially robust classification.

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

TraNce: Type-aware hypergraph neural network with biological mediators for drug repositioning.

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

A new discrimination network offers enhanced evidence-based competitive learning. This novel system improves accuracy and robustness by providing detailed unit evidence, unlike traditional competitive networks.

Area of Science:

  • Computational Neuroscience
  • Artificial Intelligence
  • Machine Learning

Background:

  • Traditional competitive networks like MAXNET primarily signal a winning unit.
  • Limited information is provided regarding the evidence supporting the winning unit's selection.

Purpose of the Study:

  • To introduce a novel discrimination network architecture for maximum selection.
  • To enhance competitive learning by incorporating detailed evidence from differentiating units.

Main Methods:

  • Construction of a discrimination network utilizing differentiating units.
  • Analysis of network convergence properties and derivation of key characteristics.
  • Incorporation of distributed redundancy to improve robustness and accuracy.

Related Experiment Videos

Main Results:

  • The discrimination network converges to a stable state in finite time.
  • Three key characteristics were derived: intensity normalization (P1), contrast enhancement (P2), and evidential response (P3).
  • Distributed redundancy enhances robustness against unit failure and noisy data, sharpening the evidential response.

Conclusions:

  • The proposed discrimination network offers a significant advancement over traditional competitive architectures.
  • It provides a more informative 'evidential response' by detailing unit evidence.
  • This network serves as a connectionist model for evidence-based competitive learning.