Abstract

Near-infrared reflectance spectroscopy uses a learning algorithm to derive a set of weighting coefficients from the reflectance spectra of a reference sample set. These coefficients, when applied to the reflectance values of an unknown sample at specific wavelengths, can be used to calculate constituent concentrations. Having enough samples in the training set and enough wavelengths in the calculation procedures is essential, but there are severe drawbacks to picking too large a number. This paper describes the principles and implementation of a working procedure for objectively calculating the minimum number of training samples required.

Evaluating logarithmic kernel for spectral reflectance estimation—effects on model parametrization, training set size, and number of sensor spectral channels

Timo Eckhard, Eva M. Valero, Javier Hernández-Andrés, and Ville Heikkinen
J. Opt. Soc. Am. A 31(3) 541-549 (2014)

Optimal hemoglobin extinction coefficient data set for near-infrared spectroscopy

Yue Zhao, Lina Qiu, Yunlong Sun, Chong Huang, and Ting Li
Biomed. Opt. Express 8(11) 5151-5159 (2017)

Dataset artificial augmentation with a small number of training samples for reflectance estimation

Jingjing Zhang, Zewei Wang, and Yuke He
Opt. Express 31(5) 8005-8019 (2023)

Nondestructive determination of SSC in an apple by using a portable near-infrared spectroscopy system

Yizhe Zhang, Jipeng Huang, Qiulei Zhang, Jinwei Liu, Yanli Meng, and Yan Yu
Appl. Opt. 61(12) 3419-3428 (2022)

Adaptive global training set selection for spectral estimation of printed inks using reflectance modeling

Timo Eckhard, Eva M. Valero, Javier Hernández-Andrés, and Markus Schnitzlein
Appl. Opt. 53(4) 709-719 (2014)