Ensemble Multivariate Calibration Based on Mutual Information for Food Analysis Using Near-Infrared Spectroscopy

Abstract

An ensemble multivariate calibration algorithm, termed as MISEPLS, is proposed. In MISEPLS, when constructing a member model, the variables that have mutual information (MI) with the response less than a threshold are eliminated; thus, the modeling can be performed in a subset of original variables and some problems arising from multi-collinearity can be avoided. Through experiments on three near-infrared (NIR) spectroscopic datasets from the food industry, MISEPLS proves to be superior to the single-model full-spectrum PLS and MIPLS (PLS combined with MI-induced variable selection). MISEPLS can improve the accuracy and robustness of a calibration model, without increasing its complexity.

Detection of Adulteration in Infant Formula Based on Ensemble Convolutional Neural Network and Near-Infrared Spectroscopy

“…Ensemble learning, which combines a number of base estimators to achieve better predictive performance [ 38 , 39 , 40 ], is an appropriate candidate for addressing the overfitting and random initialization problems of CNN. In the former studies, most of the ensemble learning approaches for NIR spectra processing adopted PLS, which is a classical linear regressor, as the base estimators [ 41 , 42 , 43 ].…”