摘要
本文使用一种新的多元统计建模方法：主成分——逐步回归法，用于近红外光谱定量分析中，其预测精度接近偏最小二乘法．
If Near-Infrared spectra (NIR) is used,to create a math model through chemometrics is neededIn this study,we put forward Principal componentstepwise regression (PCR-SRA) algorithmIn this way,the principal components are created by PCR firstThen,in term of every output variable the independent variables be selected and the regression equations is constructed each otherIn prediction,this way can compare beauty with PLS
出处
《北京农学院学报》
1999年第2期45-49,共5页
Journal of Beijing University of Agriculture
关键词
近红外光谱
定量分析
主成分
逐步回归法
Near-Infrared spectra,quantitative Analysis,Principal componentstepwise regression algorithm
