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

Automatic kernel regression modelling using combined leave-one-out test score and regularised orthogonal least

X Hong1, S Chen, P M Sharkey

  • 1Department of Cybernetics, University of Reading, Reading, RG6 6AY, UK. x.hong@cyber.reading.ac.uk

International Journal of Neural Systems
|March 23, 2004
PubMed
Summary

This study presents an automated algorithm for robust nonlinear identification using the Predicted REsidual Sums of Squares (PRESS) statistic and regularized orthogonal least squares. It enhances model robustness and generalization for accurate system identification.

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 robust nonlinear identification algorithm using PRESS statistic and forward regression.

IEEE transactions on neural networks·2008
Same author

Interleukin-1 beta regulates the expression of a leukocyte type of 12-lipoxygenase in rat islets and RIN m5F cells.

Endocrinology·1995
Same author

Mechanism of GroEL action: productive release of polypeptide from a sequestered position under GroES.

Cell·1995
Same author

Social support among impoverished women.

Nursing research·1995
Same author

Assessment of antiretroviral therapy by plasma viral load testing: standard and ICD HIV-1 p24 antigen and viral RNA (QC-PCR) assays compared.

Journal of acquired immune deficiency syndromes and human retrovirology : official publication of the International Retrovirology Association·1995
Same author

Regulation of the activity and phosphorylation of the plasma membrane Ca(2+)-ATPase by adriamycin in intact human erythrocytes.

Archives of biochemistry and biophysics·1995

Area of Science:

  • Engineering
  • Computer Science

Background:

  • Nonlinear system identification is crucial for understanding complex dynamics.
  • Existing methods often require manual tuning or separate validation datasets.

Purpose of the Study:

  • To develop an automated and robust nonlinear identification algorithm.
  • To enhance model generalization and minimize reliance on external validation data.

Main Methods:

  • Utilizing the Predicted REsidual Sums of Squares (PRESS) statistic.
  • Implementing regularized orthogonal least squares with ridge regression.
  • Developing a recursive computation formula for PRESS errors.

Main Results:

  • Successfully derived PRESS error for regularized orthogonal weight models.

Related Experiment Videos

  • Established an efficient recursive computation for PRESS errors.
  • Constructed models with improved generalization properties.

    • Conclusions:

    • The proposed algorithm enables fully automated model construction.
    • Achieves maximized model robustness through parameter regularization and structure selection.
    • Eliminates the need for separate validation datasets for model evaluation.