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

The loading problem for recursive neural networks.

Marco Gori1, Alessandro Sperduti

  • 1Dipartimento di Ingegneria dell'Informazione, Università di Siena, Via Roma 56, Siena 53100, Italy. marco@dii.unisi.it

Neural Networks : the Official Journal of the International Neural Network Society
|October 4, 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

MolVE: An Open-Source Web Platform for Visualizing and Evaluating AI-Designed Molecules to Aid in Prioritization.

Journal of chemical information and modeling·2026
Same author

Artificial intelligence in medicine: a position paper by the Italian Society of Internal Medicine.

Internal and emergency medicine·2025
Same author

A systematic literature review of spatio-temporal graph neural network models for time series forecasting and classification.

Neural networks : the official journal of the International Neural Network Society·2025
Same author

Green design in living and bedroom spaces: exploring environmental restorativeness and affective qualities of spaces.

Frontiers in psychology·2025
Same author

Interoperable Traceability in Agrifood Supply Chains: Enhancing Transport Systems Through IoT Sensor Data, Blockchain, and DataSpace.

Sensors (Basel, Switzerland)·2025
Same author

On the application of neural networks for structured domains to fMRI data.

Neural networks : the official journal of the International Neural Network Society·2025
Same journal

Exploiting audio-visual modalities in videos: Object detection via multi-stage bilateral coupling network.

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

Reliability-aware modality completion with cross-modal distillation for federated learning with missing modalities.

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

IGFD-Net: Illumination-guided frequency decoupling for polarization image fusion.

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

Multiple-Strategies dung beetle optimizer and its applications in engineering optimization and bankruptcy prediction.

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

Aggregating global-scale pixel-wise forgery cues within a graph.

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

Finite-Time intermittent control for secure synchronization of Neutral-Type stochastic delayed neural networks under aperiodic DoS attacks.

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

This study explores how neural network architecture impacts learning difficulty for graphical data. It introduces topological indices to design recursive neural networks for directed acyclic graphs without local minima in the error function.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Supervised connectionist models, particularly neural networks, face challenges in understanding the relationship between task difficulty and architecture.
  • Existing research primarily focuses on vector and sequence processing, with limited studies on graphical inputs.
  • The presence of local minima in the error function complicates the training of neural networks.

Purpose of the Study:

  • To investigate the relationship between learning task difficulty and neural network architecture for graphical inputs.
  • To establish conditions guaranteeing the absence of local minima in the error function for learning directed acyclic graphs.
  • To introduce novel topological indices for designing local minima-free neural network architectures.

Main Methods:

Related Experiment Videos

  • Developing sufficient conditions for the absence of local minima in error functions for directed acyclic graph learning.
  • Introducing topological indices calculable from training data to guide neural architecture design.
  • Conceiving a reduction algorithm incorporating node information and topology to ensure unimodal error functions.

Main Results:

  • Sufficient conditions are presented for guaranteeing an absence of local minima when learning directed acyclic graphs with recursive neural networks.
  • Novel topological indices are introduced, enabling the design of neural architectures with local minima-free error functions.
  • A reduction algorithm is developed that enhances the class of problems with unimodal error functions.

Conclusions:

  • The proposed topological indices and reduction algorithm offer a method for designing effective neural networks for graphical data processing.
  • This work contributes to a deeper understanding of supervised connectionist models and their application to complex graph-based learning tasks.
  • The findings facilitate the development of more robust and efficient training processes for neural networks handling graph structures.