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

A training algorithm for binary feedforward neural networks.

D L Gray1, A N Michel

  • 1Dept. of Electr. Eng., Purdue Univ., Hammond, IN.

IEEE Transactions on Neural Networks
|January 1, 1992
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

Effects of dietary N-(4-hydroxyphenyl)retinamide on N-nitrosomethylbenzylamine metabolism and esophageal tumorigenesis in the Fischer 344 rat.

Journal of the National Cancer Institute·2001
Same author

The effects of a binary mixture of benzo(a)pyrene and 7H-dibenzo(c,g)carbazole on lung tumors and K-ras oncogene mutations in strain A/J mice.

Experimental lung research·2001
Same author

Abdominal circumference vs. estimated weight to predict large for gestational age birth weight in diabetic pregnancy.

Clinical imaging·2000
Same author

Sensitivity and specificity of ultrasound for the detection of neural tube and ventral wall defects in a high-risk population.

Obstetrics and gynecology·1999
Same author

Predictors of persistence of adnexal masses in pregnancy.

Obstetrics and gynecology·1999
Same author

Interaction of the anxiogenic agent, RS-30199, with 5-HT1A receptors: modulation of sexual activity in the male rat.

Neuropharmacology·1998
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

A novel Boolean-like training algorithm (BLTA) significantly reduces training times for four-layer binary feedforward neural networks (BFNNs). This method enhances data generalization for Boolean functions and can process real-valued inputs with analog hardware.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Feedforward neural networks are widely used for various computational tasks.
  • Traditional training algorithms can be computationally intensive and time-consuming.
  • Boolean algebra principles offer a unique approach to neural network training.

Purpose of the Study:

  • To introduce a new training algorithm for binary-to-binary mappings using a four-layer perceptron-type feedforward neural network.
  • To develop an algorithm derived from Boolean algebra principles for efficient neural network training.
  • To enable generalization of data for incompletely specified Boolean functions.

Main Methods:

  • The Boolean-like training algorithm (BLTA) was developed based on Boolean algebra principles with extensions.

Related Experiment Videos

  • The algorithm is implemented on a four-layer binary feedforward neural network (BFNN).
  • The BFNN can be augmented with Analog-to-Digital (A/D) converters to handle real-valued inputs.

    • Main Results:

    • The BLTA demonstrates significantly reduced training times compared to descent-based methods.
    • The algorithm facilitates generalization of data, even for incompletely specified Boolean functions.
    • The BFNN with A/D converters extends the algorithm's applicability to real-valued inputs.

    Conclusions:

    • The BLTA offers a more efficient training method for binary feedforward neural networks.
    • This approach provides enhanced capabilities for handling Boolean functions and real-valued data.
    • The BLTA represents a novel technique in neural network training, distinct from traditional circuit building.