Cost-efficient unsupervised sample selection for multivariate calibration

Indirect quantification of chemical composition through spectral measurements requires the establishment of multivariate calibration models. The reference analyses on the calibration samples typically form a major cost factor in the establishment of these multivariate models. Therefore, the aim of this study was to select the most informative calibration samples in an unsupervised way based on the spectral measurements. To this end, guidelines to address this challenge in PLSR model building have been developed. The recommendations include calculating a sample size that surpasses the model complexity by a factor of 12, performing the selection in the PCA score space spanned by a sufficiently large number of principal components and using methods such as Kennard-Stone, Puchwein, Clustering or D-optimal designs. We provide the data and methodology used in the present study for future use.

Automated summary

This study aims to select the most informative calibration samples in an unsupervised way based on spectral measurements by providing guidelines for addressing challenges in PLSR model building. Recommendations include calculating a sample size exceeding the model complexity by a factor of 12, performing selection in a PCA score space with a sufficient number of principal components, and using methods such as Kennard-Stone.