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

Efficient training of multilayer perceptrons using principal component analysis.

Christoph Bunzmann1, Michael Biehl, Robert Urbanczik

  • 1Institut für Theoretische Physik, Universität Würzburg Am Hubland, D-97074 Würzburg, Germany.

Physical Review. E, Statistical, Nonlinear, and Soft Matter Physics
|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

Enhanced broad-band intermuscular coherence in myoclonus: a targeted characterization study.

Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology·2026
Same author

Machine learning based EMG analysis of intermuscular coherence and cumulant density in tremor and myoclonus.

Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology·2026
Same author

Temporal stimulus segmentation by reinforcement learning in populations of spiking neurons.

Physical review. E·2026
Same author

Endocrine and metabolic determinants of cardiometabolic risk in mild autonomous cortisol secretion.

EBioMedicine·2026
Same author

Explainable machine learning for movement disorders - Classification of tremor and myoclonus.

Computers in biology and medicine·2025
Same author

Subspace corrected relevance learning with application in neuroimaging.

Artificial intelligence in medicine·2024

This study introduces a novel training algorithm for multilayer perceptrons, enhancing generalization with principal component analysis. The method requires significantly fewer training examples than traditional approaches for effective machine learning.

Area of Science:

  • Machine Learning
  • Statistical Physics
  • Artificial Neural Networks

Background:

  • Multilayer perceptrons (MLPs) are fundamental in machine learning.
  • Traditional online training methods for MLPs can be data-intensive and prone to overfitting.
  • Principal Component Analysis (PCA) is a dimensionality reduction technique with potential applications in optimizing learning algorithms.

Purpose of the Study:

  • To introduce and analyze a novel training algorithm for MLPs.
  • To investigate the relationship between this algorithm and principal component analysis (PCA).
  • To compare the generalization performance of the new algorithm against traditional online training methods.

Main Methods:

  • Developing a training algorithm for MLPs incorporating PCA.

Related Experiment Videos

  • Performing statistical physics analysis on learning models for regression and classification tasks.
  • Evaluating the algorithm's properties using theoretical models and simulations.

    • Main Results:

    • The proposed training algorithm demonstrates superior generalization capabilities compared to traditional online training.
    • The algorithm requires substantially fewer training examples for effective learning.
    • An optimal training prescription for networks with numerous hidden units was derived, maximizing generalization within the model.

    Conclusions:

    • The novel MLP training algorithm offers significant advantages in terms of data efficiency and generalization.
    • This approach, leveraging PCA, provides a powerful alternative for training complex neural networks.
    • The findings suggest a new direction for developing more efficient and effective machine learning algorithms.