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

Learning and generalization in cascade network architectures

E Littmann1, H Ritter

  • 1Abt. Neuroinformatik, Fakultät für Informatik, Universität Ulm, Germany.

Neural Computation
|October 1, 1996
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

High adherence to anti-tuberculosis treatment among patients attending a hospital and slum health centre in Nairobi, Kenya.

Global public health·2024
Same author

A helium-burning white dwarf binary as a supersoft X-ray source.

Nature·2023
Same author

Canadian Cancer Trials Group (CCTG) IND215: A phase Ib study of Selumetinib in patients with untreated advanced or metastatic NSCLC who are receiving standard chemotherapy regimens.

Investigational new drugs·2018
Same author

Cobalt-containing bioactive glasses reduce human mesenchymal stem cell chondrogenic differentiation despite HIF-1α stabilisation.

Journal of the European Ceramic Society·2018
Same author

Experimental amine-epoxide sealer: a physicochemical study in comparison with AH Plus and EasySeal.

International endodontic journal·2014
Same author

Crassulacean acid metabolism (CAM) in leaves of Aloe arborescens mill : Comparative studies of the carbon metabolism of chlorenchym and central hydrenchym.

Planta·2013
Same journal

A Model-Free Reinforcement Learning Implementation of Decision Making Under Uncertainty by Sequential Sampling.

Neural computation·2026
Same journal

DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning.

Neural computation·2026
Same journal

Hierarchical Active Inference Using Successor Representations.

Neural computation·2026
Same journal

W-Kernel and Its Principal Space for Frequentist Evaluation of Bayesian Estimators.

Neural computation·2026
Same journal

A Hidden Markov Model-Inspired Sequence Classification Method for Hyperdimensional Computing.

Neural computation·2026
Same journal

Sparse Graphical Modeling for Electrophysiological Phase-Based Connectivity Using Circular Statistics.

Neural computation·2026
See all related articles

Direct Cascade Architecture (DCA) enhances network performance by cascading subnetworks, even those without error-backpropagation. This approach improves generalization on small datasets, outperforming shallow architectures.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Neural Networks

Background:

  • Incrementally constructed cascade architectures offer an alternative to fixed-size networks.
  • Existing methods include the cascade-correlation approach and direct cascade architecture (DCA).

Purpose of the Study:

  • To compare Direct Cascade Architecture (DCA) with cascade-correlation and related methods.
  • To evaluate the properties and performance of DCA using benchmark results.
  • To demonstrate the enhancement of network cascades by using Large Language Model (LLM) networks as cascaded elements.

Main Methods:

  • Comparison of DCA with cascade-correlation and other incremental network construction approaches.
  • Evaluation of network properties and performance through benchmark datasets.

Related Experiment Videos

  • Application of DCA using LLM networks as subnetworks for enhanced performance.

    • Main Results:

    • DCA allows cascading of entire subnetworks, including those without error-backpropagation.
    • Cascaded LLM networks significantly enhance performance compared to single networks.
    • Deeply cascaded DCA architectures show good generalization on small datasets, avoiding overfitting.

    Conclusions:

    • DCA provides a powerful and flexible method for constructing modular systems.
    • It is a viable alternative to existing schemes like mixtures of experts.
    • DCA facilitates the integration of diverse subnetwork types into modular systems.