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 comparison of non-linear non-parametric models for epilepsy data.

F Miwakeichi1, R Ramirez-Padron, P A Valdes-Sosa

  • 1The Graduate University for Advanced Studies, 4-6-7 Minami Azabu, Minato-ku 106-0047, Tokyo, Japan. miwake1@ism.ac.jp

Computers in Biology and Medicine
|November 4, 2000
PubMed
Summary

The Nadaraya-Watson (NW) method effectively models EEG spike and wave (SW) activity, outperforming local linear regression and SVMs, especially with added noise. This stochastic modeling approach enhances SW reproduction accuracy.

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

Publisher Correction: Brain charts for the human lifespan.

Nature·2022
Same author

Brain charts for the human lifespan.

Nature·2022
Same author

Spatio-temporal correlations from fMRI time series based on the NN-ARx model.

Journal of integrative neuroscience·2011
Same author

The relationship between the visually evoked P300 event-related potential and gamma band oscillation in the human medial and basal temporal lobes: an electrocorticographic study.

Neuroscience research·2002
Same author

The influence of craniofacial form on bite force and EMG activity of masticatory muscles. VIII-1. Bite force of complete denture wearers.

Journal of oral science·1999
Same author

Utility of early second-look laparoscopy after infertility surgery.

The Journal of the American Association of Gynecologic Laparoscopists·1999

Area of Science:

  • Neuroscience
  • Computational Biology
  • Signal Processing

Background:

  • EEG spike and wave (SW) activity is a key biomarker in neurological disorders.
  • Non-parametric stochastic models offer a powerful framework for analyzing complex biological signals like SW.
  • Previous models have limitations in accurately reproducing SW, particularly under noisy conditions.

Purpose of the Study:

  • To compare the performance of Nadaraya-Watson (NW), local linear polynomial regression, and Support Vector Machines (SVM) for modeling EEG SW activity.
  • To evaluate the robustness of these methods when incorporating dynamical noise.
  • To assess the accuracy of noise-free SW realizations generated by different modeling techniques.

Main Methods:

  • Non-parametric stochastic modeling using Nadaraya-Watson (NW) regression.

Related Experiment Videos

  • Local linear polynomial regression.
  • Support Vector Machines (SVM).
  • Estimation of tuning parameters.
  • Introduction of dynamical noise.
  • Correlation dimension analysis for noise standard deviation estimation.

    • Main Results:

    • NW and SVM methods produced improved noise-free SW realizations compared to prior studies.
    • Manual estimation of tuning parameters was required for NW and SVM.
    • Only the NW method successfully generated SW similar to training data when dynamical noise was added.
    • Correlation dimension effectively estimated the standard deviation of dynamical noise.

    Conclusions:

    • The Nadaraya-Watson (NW) method demonstrates superior performance in modeling EEG spike and wave (SW) activity, particularly in the presence of dynamical noise.
    • NW and SVM offer enhanced accuracy for generating noise-free SW realizations.
    • The NW method's ability to handle noise makes it a promising tool for analyzing complex neurological signals.