Multiblock partial least squares regression based on wavelet transform for quantitative analysis of near infrared spectra
Multiblock partial least squares regression based on wavelet transform for quantitative analysis of near infrared spectra
复制标题
基于小波变换的多块偏最小二乘回归近红外光谱定量分析
DOI：
10.1016/j.chemolab.2009.09.006
复制
发表时间：
2010-01-15
影响因子：
3.9
通讯作者：
Shao, Xueguang
中科院分区：
文献类型：
--
作者：
Jing, Ming;Cai, Wensheng;Shao, Xueguang
多块偏最小二乘（MB-PLS）方法已被提出来建模的数据集与大量的变量，使模型更易于解释。在MB-PLS中，变量被分成包含不同信息的若干块，并且块的相对重要性由MB-PLS模型的超权重来反映。本文提出了一种加权MB-PLS与离散小波变换（DWT）耦合的近红外光谱建模方法。该方法利用小波变换对光谱进行分块处理，通过交叉验证子模型预测误差确定的块权值和超权值来估计块的相对重要性。因此。为MB-PLS提供了一种分离变量的实用方法，并且可以自适应地调制可变块对预测的相对贡献。验证该方法的性能。分别研究了烟草粉末和烟草薄片碎片的两组工业近红外数据。均方根预测误差（RMSEP）。残差预测偏差（RPD）和相关系数（R）表明，加权MB-PLS与DWT耦合的预测精度和可解释性优于普通PLS和MB-PLS方法。(C)2009 Elsevier B. V.保留所有权利。
Multiblock partial least squares (MB-PLS) method has been proposed for modeling the data set with large number of variables and for making the model more interpretable. In MB-PLS, the variables are split into several blocks containing different information, and the relative importance of the blocks is reflected by the super-weights of the MB-PLS model. In this paper, a weighted MB-PLS Coupled with discrete wavelet transform (DWT) method is proposed for modeling of the near infrared (NIR) spectra. In the method, the spectra are decomposed into blocks by DWT, and the relative importance of the blocks is estimated by both the super-weights and the block-weights determined by the prediction error of the sub-models in cross validation. Therefore. a practical approach to separate the variables is provided for MB-PLS and the relative contribution of the variable blocks to the prediction can be modulated adaptively. To validate the performance of the method. two industrial NIR data sets of tobacco powder and fragment of tobacco lamina are investigated, respectively. The root-mean-square error of prediction (RMSEP). the residual predictive deviation (RPD), and the correlation coefficient (R) show that the weighted MB-PLS Coupled with DWT gives a better predictive accuracy and interpretability compared with the ordinary PLS and MB-PLS methods. (C) 2009 Elsevier B.V. All rights reserved.