Linear model correction: A method for transferring a near-infrared multivariate calibration model without standard...

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Chemometrics

Background:

• Calibration transfer is crucial for applying near-infrared (NIR) spectroscopy across different instruments.
• Existing methods often require standard samples to correct for instrument variations, limiting practical use.

Purpose of the Study:

• To propose a novel calibration transfer method that eliminates the need for standard samples.
• To develop a technique for accurate spectral data prediction between dissimilar NIR instruments.

Main Methods:

• Linear Model Correction (LMC) method based on linear correlation between spectra from different instruments.
• Constrained optimization to transfer model coefficients from a master to a slave instrument.
• Validation using NIR datasets of corn and plant leaf samples.

Main Results:

• The LMC method successfully transferred partial least squares (PLS) models between instruments.
• Accurate spectral predictions were achieved for both corn and plant leaf datasets.
• The method demonstrated effectiveness without the requirement of standard samples.

Conclusions:

• The proposed LMC method offers a practical solution for calibration transfer in NIR spectroscopy.
• Eliminating the need for standard samples enhances the method's applicability in real-world scenarios.
• This technique facilitates more widespread and efficient use of NIR spectroscopy.