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

Least-squares methods for blind source separation based on nonlinear PCA.

P Pajunen1, J Karhunen

  • 1Helsinki University of Technology, Laboratory of Computer and Information Science, Espoo, Finland. Petteri.Pajunen@hut.fi

International Journal of Neural Systems
|March 5, 1999
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

Physically principled reflection models applied to filtered camera imaging inversions in metal walled fusion machines.

The Review of scientific instruments·2019
Same author

Characterising metabolically healthy obesity in weight-discordant monozygotic twins.

Diabetologia·2013
Same author

HbA(1c) in diagnosing and predicting Type 2 diabetes in impaired glucose tolerance: the Finnish Diabetes Prevention Study.

Diabetic medicine : a journal of the British Diabetic Association·2010
Same author

Impact of early angiographic evaluation on the frequency of emergency reoperations after coronary bypass surgery.

Scandinavian journal of surgery : SJS : official organ for the Finnish Surgical Society and the Scandinavian Surgical Society·2010
Same author

The metabolic syndrome as a predictor of incident diabetes and cardiovascular events in the Health 2000 Study.

Diabetes & metabolism·2010
Same author

A European evidence-based guideline for the prevention of type 2 diabetes.

Hormone and metabolic research = Hormon- und Stoffwechselforschung = Hormones et metabolisme·2010

This study introduces novel least-squares algorithms for blind source separation, improving upon neural network methods. These efficient algorithms offer faster convergence for extracting signals from mixtures.

Area of Science:

  • Signal Processing
  • Machine Learning
  • Computational Neuroscience

Background:

  • Blind source separation (BSS) aims to recover original signals from mixed signals with minimal prior information.
  • Nonlinear principal component analysis (PCA) neural algorithms have shown promise for BSS.
  • Existing methods often face challenges with computational efficiency and convergence speed.

Purpose of the Study:

  • To develop computationally efficient and fast-converging algorithms for BSS.
  • To explore the application of least-squares approaches to minimize nonlinear PCA criteria for BSS.
  • To compare the performance of these new methods against existing stochastic gradient algorithms.

Main Methods:

  • Minimization of a nonlinear PCA criterion using least-squares optimization.

Related Experiment Videos

  • Development and study of several least-squares-based algorithms, including some with neural learning capabilities.
  • Investigation of the connection between least-squares methods and the nonlinear PCA subspace rule.

    • Main Results:

    • Least-squares approaches provide computationally efficient and fast-converging algorithms for BSS.
    • The developed algorithms demonstrate significant speed improvements over stochastic gradient algorithms in experimental BSS tasks.
    • A clear link is established between the least-squares minimization and the nonlinear PCA subspace rule.

    Conclusions:

    • Least-squares optimization offers a superior approach for minimizing nonlinear PCA criteria in blind source separation.
    • The proposed algorithms represent a significant advancement in BSS, offering enhanced speed and efficiency.
    • This work provides a foundation for further research into efficient signal processing techniques using nonlinear dimensionality reduction.