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 Efficient EM-based Training Algorithm for Feedforward Neural Networks.

James Farmer1, Chuanyi Ji, Sheng Ma

  • 1Rensselaer Polytechnic Institute, USA

Neural Networks : the Official Journal of the International Neural Network Society
|March 1, 1997
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

Metabolic Stability and Metabolic Profile of Schisandrol B in Rat and Human Liver Microsomes Employing HPLC-MS/MS Analysis.

Rapid communications in mass spectrometry : RCM·2026
Same author

Secondary Myxomatous Corneal Degeneration: Two Cases With Histopathologic Diagnosis and a Systematic Review Supporting Conservative Management.

Cornea·2026
Same author

Activatable Polymeric STING Agonist-Gold Nanorod Conjugate Driving STING Signaling and Immunogenic Activation in Colorectal Cancer Immunotherapy.

Nano letters·2026
Same author

Risk factors for diabetic kidney disease in type 2 diabetes mellitus in Asia: a meta-analysis.

Frontiers in endocrinology·2026
Same author

Integration of electronic nose and comprehensive two-dimensional gas chromatography-olfactometry-time-of-flight mass spectrometry to decode the olfactory quality enhancement of agarwood essential oil upon thermal treatment.

Journal of chromatography. A·2026
Same author

Effects of Voriconazole on the Safety, Pharmacokinetics and Pharmacodynamics of Ciprofol (HSK3486), a Novel Intravenous Anaesthetic, in Healthy Participants: A Prospective, Randomized, Crossover Clinical Trial.

Drug design, development and therapy·2026
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

A new algorithm speeds up training for two-layer neural networks by breaking down the process. This method trains individual neurons efficiently, significantly improving performance on benchmark tasks.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Two-layer feedforward neural networks are fundamental in machine learning.
  • Efficient training algorithms are crucial for practical applications of neural networks.
  • Existing training methods can be computationally intensive.

Purpose of the Study:

  • To develop a fast training algorithm for two-layer feedforward neural networks.
  • To leverage probabilistic models for hidden representations.
  • To improve computational efficiency in neural network training.

Main Methods:

  • Developed a novel training algorithm based on a probabilistic model for hidden representations.
  • Utilized the Expectation-Maximization (EM) algorithm.

Related Experiment Videos

  • Decomposed the training of two-layer networks into training individual neurons using linear weighted regression.

    • Main Results:

    • Achieved significant improvements in training speed.
    • Demonstrated effectiveness on several benchmark problems.
    • The algorithm decomposes complex network training into simpler, single-neuron training tasks.

    Conclusions:

    • The proposed algorithm offers a substantial speed-up for training two-layer feedforward neural networks.
    • The decomposition strategy simplifies the training process.
    • This approach enhances the practicality of using neural networks for various applications.