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

Nonlinear principal component analysis of noisy data.

William W Hsieh1

  • 1Department of Earth and Ocean Sciences, University of British Columbia, Vancouver, BC, Canada. whsieh@eos.ubc.ca

Neural Networks : the Official Journal of the International Neural Network Society
|May 22, 2007
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

Neural network forecasts of the tropical Pacific sea surface temperatures.

Neural networks : the official journal of the International Neural Network Society·2006
Same author

Complex-valued neural networks for nonlinear complex principal component analysis.

Neural networks : the official journal of the International Neural Network Society·2005
Same journal

A boundary-regularization-enhanced video anomaly detection network based on context-adaptive spatio-temporal conditional diffusion.

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

MT<sup>2</sup>-CSD and LLM-CRAN: A new dataset and an LLM-based multi-semantic knowledge fusion model for conversational stance detection.

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

TriAlignNet: A triple-path cross-modality alignment framework for multimodal time series forecasting.

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

Anchor-based disentanglement framework for incremental multi-view clustering.

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

Complex-valued amplitude-phase interference modeling for adversarially robust classification.

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

TraNce: Type-aware hypergraph neural network with biological mediators for drug repositioning.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

A new information criterion (IC) effectively selects models in nonlinear principal component analysis (NLPCA), preventing overfitting even with noisy data. This method improves model selection for complex datasets.

Area of Science:

  • Data science
  • Machine learning
  • Climate science

Background:

  • Overfitting is a challenge in nonlinear principal component analysis (NLPCA), especially with noisy data.
  • Existing methods struggle to prevent overfitting in NLPCA, unlike in nonlinear regression.
  • Model complexity and regularization are key factors in NLPCA performance.

Purpose of the Study:

  • To introduce a novel information criterion (IC) for robust model selection in NLPCA.
  • To address the issue of overfitting in NLPCA by proposing a new model selection metric.
  • To evaluate the effectiveness of the proposed IC on synthetic and real-world climate data.

Main Methods:

  • Developed a new information criterion (IC) to quantify inconsistency between nonlinear principal components (u and ũ) and their nearest neighbors.

Related Experiment Videos

  • Utilized autoassociative neural networks for performing nonlinear principal component analysis (NLPCA).
  • Applied the IC to select the optimal model from multiple NLPCA models with varying complexity and regularization.

    • Main Results:

    • The proposed IC successfully identified the best model by penalizing overfitted solutions, as indicated by increasing inconsistency (I).
    • Tests on synthetic and climate data (sea surface temperatures, stratospheric winds) demonstrated the IC's efficacy.
    • The IC effectively aided in model selection and distinguishing between open and closed curve solutions.

    Conclusions:

    • The novel information criterion provides a reliable method for model selection in nonlinear principal component analysis.
    • This approach effectively mitigates overfitting issues in NLPCA, particularly with noisy datasets.
    • The IC is a valuable tool for analyzing complex climate data and improving the interpretability of NLPCA results.