New discrimination method combining hit quality index based spectral matching and voting.

Area of Science:

• Analytical Chemistry
• Chemometrics
• Spectroscopy

Background:

• Accurate sample discrimination is crucial in various scientific fields.
• Existing multivariate methods like PCA-LDA and PLS-DA have limitations in certain discrimination tasks.
• Hit Quality Index (HQI) measures spectral matching, but its application in a voting system for enhanced discrimination was unexplored.

Purpose of the Study:

• To develop and evaluate a new discrimination method, HQI-voting, for analyzing spectral data.
• To compare the performance of HQI-voting against established multivariate techniques.
• To assess the robustness of HQI-voting in scenarios with closely related spectral features.

Main Methods:

• Developed the HQI-voting method, utilizing HQI values for discriminant analysis.
• Applied HQI-voting to three near-infrared (NIR) spectroscopic datasets: sesame, Angelica gigas, and diesel/LGO samples.
• Compared HQI-voting performance with principal component analysis-linear discriminant analysis (PCA-LDA), partial least squares-discriminant analysis (PLS-DA), and k-nearest neighbor (k-NN).

Main Results:

• HQI-voting demonstrated improved discrimination performance compared to PCA-LDA and PLS-DA across all tested datasets.
• The method showed robustness, especially when multiple library samples had similar HQI values.
• Accuracies achieved with HQI-voting were competitive with existing factor-based multivariate methods.

Conclusions:

• HQI-voting is an effective and robust method for spectral data discrimination.
• This novel approach offers comparable or superior performance to traditional multivariate techniques.
• The findings support the utility of HQI-voting as a valuable tool in chemometrics and analytical chemistry.