Enhanced 3-D asynchronous correlation data preprocessing method for Raman spectroscopy of Chinese handmade paper.

Area of Science:

• Materials Science
• Spectroscopy
• Data Science

Background:

• Accurate classification of materials like Chinese handmade paper is crucial.
• Traditional spectral analysis methods may lack sufficient resolution and hidden information extraction capabilities.

Purpose of the Study:

• To introduce and evaluate a novel 3D asynchronous correlation method (3D-ACM) for enhanced material classification.
• To assess the performance of 3D-ACM when combined with various machine learning models for spectral data analysis.

Main Methods:

• Developed a 3D-ACM involving tensor product and Hilbert transform operations on Raman spectral data.
• Applied six machine learning models (PCA-LR, SVM-LR, KNN, RF, CNN) to the preprocessed data.
• Evaluated model performance using R-squared values.

Main Results:

• 3D-ACM significantly increased spectral resolution and revealed hidden information.
• Machine learning models (PLS-LR, KNN, RF, CNN) achieved R-squared values near or equal to 1 when using 3D-ACM.
• Performance was comparable to unsupervised methods like PCA.

Conclusions:

• 3D-ACM is a versatile mathematical technique for spectral data preprocessing.
• It offers superior performance for material classification and identification without additional experimental setups.
• 3D-ACM shows significant promise for future applications in material science and data analysis.