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

Infinite-dimensional multilayer perceptrons.

M Kuzuoglu1, K Leblebicioglu

  • 1Dept. of Electr. and Electron. Eng., Middle East Tech. Univ., Ankara.

IEEE Transactions on Neural Networks
|January 1, 1996
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

A method for comparative evaluation of EIT algorithms using a standard data set.

Physiological measurement·1995
Same author

Analytic and semi-analytic solutions in electrical impedance tomography: I. Two-dimensional problems.

Physiological measurement·1995
Same author

Analytic and semi-analytic solutions in electrical impedance tomography: II. Three-dimensional problems.

Physiological measurement·1995
Same author

A fast image reconstruction algorithm for electrical impedance tomography.

Physiological measurement·1994
Same author

Electrical impedance tomography using induced currents.

IEEE transactions on medical imaging·1994
Same author

Analysis of three-dimensional software EIT (electrical impedance tomography) phantoms by the finite element method.

Clinical physics and physiological measurement : an official journal of the Hospital Physicists' Association, Deutsche Gesellschaft fur Medizinische Physik and the European Federation of Organisations for Medical Physics·1992
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 infinite-dimensional multilayer perceptron (IDMLP) simulates nonlinear transformations using integral transforms. This new structure, trained with variational techniques and Lagrange multipliers, offers enhanced function approximation capabilities.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Functional Analysis

Background:

  • Multilayer perceptrons (MLPs) are foundational in machine learning for modeling complex relationships.
  • Traditional MLPs operate on finite-dimensional spaces, limiting their application to certain complex functions.
  • Infinite-dimensional function spaces present unique challenges for neural network architectures.

Purpose of the Study:

  • To introduce a novel infinite-dimensional multilayer perceptron (IDMLP) structure.
  • To extend the capabilities of MLPs for simulating nonlinear transformations in infinite-dimensional function spaces.
  • To develop effective training methodologies for this new architecture.

Main Methods:

  • Replaced discrete neurons with a continuum of neurons.

Related Experiment Videos

  • Substituted summations with integrations and weight matrices with kernels of integral transforms.
  • Employed variational techniques and the Lagrange multiplier method for analysis and training, deriving coupled state and adjoint state integro-difference equations.
  • Utilized a steepest descent-like algorithm for constructing kernel and threshold functions.

    • Main Results:

    • Successfully developed and introduced an infinite-dimensional MLP (IDMLP).
    • Demonstrated the capability of the IDMLP to handle nonlinear transformations in infinite-dimensional function spaces.
    • Presented results showcasing the performance and efficacy of the proposed IDMLP structure.

    Conclusions:

    • The proposed IDMLP offers a powerful new framework for modeling complex nonlinearities in infinite-dimensional spaces.
    • The integration of variational techniques and integral transforms provides a robust method for training and analysis.
    • The IDMLP shows significant potential for advancing machine learning applications dealing with function spaces.