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

Wavelength selection for multivariate calibration using tikhonov regularization.

Forrest Stout1, John H Kalivas, Károly Héberger

  • 1Department of Chemistry, Idaho State University, Pocatello, Idaho 83209, USA.

Applied Spectroscopy
|February 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

Sum of Euclidean distance differences and sum of absolute Manhattan distance differences: multicriteria decision making tools for small data tables.

Analytica chimica acta·2025
Same author

Perspective on the Capacity of the Rashomon Effect in Multivariate Data Analysis.

Applied spectroscopy·2025
Same author

Local Adaptive Fusion Regression (LAFR) for Local Linear Multivariate Calibration: Application to Large Datasets.

Applied spectroscopy·2025
Same author

Alternative weighting schemes for fine-tuned extended similarity indices.

Journal of chemometrics·2024
Same author

Redefining Spectral Data Analysis with Immersive Analytics: Exploring Domain-Shifted Model Spaces for Optimal Model Selection.

Applied spectroscopy·2024
Same author

Local Modeling by Adapting Source Calibration Models to Analyte Shifted Target Domain Samples Without Reference Values.

Applied spectroscopy·2024

This study introduces Tikhonov regularization with the 1-norm (TR1) for built-in wavelength selection in spectroscopic analysis. TR1 models effectively reduce coefficients for irrelevant wavelengths, improving prediction accuracy in multivariate calibration.

Area of Science:

  • Chemometrics
  • Spectroscopy
  • Data Analysis

Background:

  • Wavelength selection enhances prediction accuracy in spectroscopic data analysis.
  • Traditional methods like stepwise regression (SWR) are commonly used.
  • Evaluating bias/variance tradeoffs and parsimony is crucial for model selection.

Purpose of the Study:

  • To introduce and evaluate a novel built-in wavelength selection method based on Tikhonov regularization with the 1-norm (TR1).
  • To compare TR1 with stepwise regression (SWR) and full-wavelength models (PCR, RR, PLS, MLR).
  • To provide guidelines on selecting appropriate wavelength strategies for spectroscopic calibration.

Main Methods:

  • Application of Tikhonov regularization with the 1-norm (TR1) for automatic wavelength selection.

Related Experiment Videos

  • Comparison with stepwise regression (SWR) on simulated and near-infrared (NIR) spectral data.
  • Analysis of bias, variance, and parsimony for various multivariate calibration techniques (PCR, RR, PLS, MLR) with and without wavelength selection.

    • Main Results:

    • TR1 models inherently perform wavelength selection by assigning near-zero coefficients to undesirable wavelengths.
    • TR1 models often outperform full-wavelength PCR, RR, and PLS models in prediction accuracy.
    • SWR subset results were comparable to TR1 for NIR data but inferior for simulated data.
    • Wavelength selection generally improves prediction accuracy but may slightly increase variance.

    Conclusions:

    • TR1 offers an effective built-in wavelength selection approach for spectroscopic calibration.
    • The choice between full wavelengths and subset selection depends on the nature and number of spectral effects.
    • TR1 is particularly advantageous when dealing with spectral interferences or poor signal-to-noise ratios.