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 regularized kernel CCA contrast function for ICA.

Carlos Alzate1, Johan A K Suykens

  • 1Department of Electrical Engineering ESAT-SCD-SISTA, Katholieke Universiteit Leuven, B-3001 Leuven, Belgium. carlos.alzate@esat.kuleuven.be

Neural Networks : the Official Journal of the International Neural Network Society
|February 19, 2008
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

Deep Kernel Principal Component Analysis for multi-level feature learning.

Neural networks : the official journal of the International Neural Network Society·2023
Same author

Learning With Asymmetric Kernels: Least Squares and Feature Interpretation.

IEEE transactions on pattern analysis and machine intelligence·2023
Same author

Disentangled Representation Learning and Generation With Manifold Optimization.

Neural computation·2022
Same author

Compressing Features for Learning With Noisy Labels.

IEEE transactions on neural networks and learning systems·2022
Same author

Towards a Unified Quadrature Framework for Large-Scale Kernel Machines.

IEEE transactions on pattern analysis and machine intelligence·2021
Same author

Unsupervised learning of disentangled representations in deep restricted kernel machines with orthogonality constraints.

Neural networks : the official journal of the International Neural Network Society·2021
Same journal

Dynamic analysis and reliable mechanical optimization application of ring HNN effected with a memristive neuron.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

DAFF-Net: A detection and search method for small-scale low surface brightness galaxies.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Quasi-synchronization for complex networks with hybrid pinning intermittent control.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Physics-encoded convolutional neural operators for parametric PDEs: A convergence-guaranteed framework via pre-computed kernel fields.

Neural networks : the official journal of the International Neural Network Society·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
See all related articles

A novel kernel-based method enhances independent component analysis (ICA) by introducing a regularized correlation measure. This approach improves component estimation accuracy, particularly for complex datasets like images and speech signals.

Area of Science:

  • Machine Learning
  • Signal Processing
  • Statistical Analysis

Background:

  • Independent Component Analysis (ICA) is crucial for signal separation.
  • Existing kernel-based methods for ICA have limitations in high-dimensional spaces.
  • Canonical Correlation Analysis (CCA) provides a framework for relating variables.

Purpose of the Study:

  • To propose a new kernel-based contrast function for ICA.
  • To extend Least Squares Support Vector Machine (LS-SVM) and CCA formulations to kernel ICA.
  • To develop a statistically reliable measure of independence applicable to out-of-sample data.

Main Methods:

  • Formulation of a regularized correlation measure in kernel-induced feature spaces.
  • Extension of LS-SVM to kernel CCA, incorporating regularization in the primal problem.

Related Experiment Videos

  • Solving the dual generalized eigenvalue problem to find the smallest eigenvalue as a measure of independence.
  • Utilizing incomplete Cholesky factorization to address computational challenges in eigendecomposition.

    • Main Results:

    • The smallest generalized eigenvalue effectively measures correlation in feature space and independence in input space.
    • The primal-dual approach allows for reliable out-of-sample extension of the independence measure.
    • Simulations demonstrate superior performance in estimating independent components compared to existing kernel-based contrast functions.
    • Improved results were observed across toy data, image, and speech signal processing tasks.

    Conclusions:

    • The proposed kernel-based contrast function offers a robust and effective approach to ICA.
    • The method provides a statistically reliable measure of independence with out-of-sample capabilities.
    • The technique shows significant performance improvements over existing kernel-based ICA methods, particularly in complex signal processing applications.