Multiple regression systems for spectrophotometric data analysis

Multiple architecture system for wind speed prediction

“…It has been shown that the average strategy could be more efficient if all predictors are unbiased and have uncorrelated errors with similar variances or, in general, when all the single predictors reveal comparable accuracies [41,42].…”

NIR and PLS discriminant analysis for predicting the processability of malt during lautering

“…In addition, the method for classification should be compared to other approaches presented in the literature (such as combination of support vector machines (SVM) and particle swarm optimization (PSO), Melgani and Bazi [38] or tabu search, PSO and SVM, Chuang et al [39]). Furthermore, spectroscopic data treatment could be enhanced by multiple regression systems (see Benoudjit et al [40]). It has to be mentioned that the studies on variable selection and outlier detection could also contribute to stability and improvement of models.…”

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

“…Note that, if the pseudo inverse matrix F results ill-conditioned, one could resort to singular value decomposition (SVD) so that to avoid the matrix inversion operation (Benoudjit, Melgani, and Bouzgou 2009). As a particular case of WAF, if the same weight (1=M) is assigned to all models, there is no need to calculate these weight values.…”