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 with Minimerror, a temperature-dependent learning algorithm

B Raffin1, M B Gordon

  • 1CEA/Département de Recherche Fondamentale sur la Matière Condensée, SPSMS/MDN, Centre d'Etudes Nucléaires de Grenoble, France.

Neural Computation
|November 1, 1995
PubMed
Summary

Minimerror, a new perceptron learning algorithm, demonstrates optimal performance in numerical simulations for linearly separable boolean functions. Its implementation aligns perfectly with theoretical predictions, validating its efficiency.

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

Correction to: Oral octreotide capsules for the treatment of acromegaly: comparison of 2 phase 3 trial results.

Pituitary·2021
Same author

Oral octreotide capsules for the treatment of acromegaly: comparison of 2 phase 3 trial results.

Pituitary·2021
Same author

mRNA of luteal genes associated with progesterone synthesis, maintenance, and apoptosis in dairy heifers and lactating dairy cows.

Animal reproduction science·2010
Same author

Pregnancy rates to timed artificial insemination in dairy cows treated with gonadotropin-releasing hormone or porcine luteinizing hormone.

Theriogenology·2009
Same author

Variations in adequate levothyroxine replacement therapy in patients with different causes of hypothyroidism.

Endocrine practice : official journal of the American College of Endocrinology and the American Association of Clinical Endocrinologists·2004
Same author

SpikeCell: a deterministic spiking neuron.

Neural networks : the official journal of the International Neural Network Society·2003

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Computational Neuroscience

Background:

  • The perceptron algorithm is fundamental in machine learning for classification tasks.
  • Optimality in learning algorithms is crucial for efficient pattern recognition.
  • Minimerror offers theoretical advantages for both linear and nonlinear separability.

Purpose of the Study:

  • To evaluate the numerical performance of the Minimerror learning algorithm.
  • To implement and test Minimerror on linearly separable boolean functions.
  • To compare numerical results with theoretical predictions for Minimerror's efficiency.

Main Methods:

  • Implementation of the Minimerror algorithm.
  • Numerical simulations using linearly separable boolean functions.

Related Experiment Videos

  • Analysis of algorithm performance metrics.

    • Main Results:

    • Numerical results confirm the optimal performance of Minimerror.
    • The implemented algorithm achieved results consistent with theoretical predictions.
    • Demonstrated efficiency in learning linearly separable boolean functions.

    Conclusions:

    • Minimerror is a highly effective algorithm for learning linearly separable boolean functions.
    • Numerical validation supports the theoretical optimality of Minimerror.
    • The study confirms Minimerror's practical applicability in machine learning.