文献类型：会议论文

英文题名：Laplacian regularized least squares regression and its dynamic parameter optimization for near infrared spectroscopy modeling

作者：Yang, Hui-hua[1,2];Qin, Feng[1];Wang, Yong[2,3];Liang, Qiong-lin[2];Wang, Yi-ming[2];Luo, Guo-an[2]

机构：[1]Tsinghua Univ, Anal Ctr, Beijing 100084, Peoples R China;[2]Tsinghua Univ, Anal Ctr, Beijing 100084, Peoples R China;[3]East China Univ Sci & Technol, Sch Pharm, Modern Engn Ctr Tradit Chinese Med, Shanghai 200237, Peoples R China

会议论文集：3rd International Conference on Natural Computation (ICNC 2007)

会议日期：AUG 24-27, 2007

会议地点：Haikou, PEOPLES R CHINA

语种：英文

摘要：Partial Least Square (PLS) is the most commonly used algorithm for Near Infrared (NIR) modeling. NIR modeling features that it cheap, easy and fast to measure the NIR spectroscopy, while expensive, difficult and time- consuming to measure the reference value for this spectroscopy. PLS often faces the challenge of that limited samples are available in training set to build a predicative model. To tackle this problem, a novel NIR modeling method-Laplacian Regularized Least Squares Regression (LapRLSR) and its dynamically adaptive parameters optimization method was presented. Based on the semi-supervised learning framework, LapRLSR can take the advantage of many unlabeled spectra to promote the prediction performance of the model though there are only few labeled samples. The proposed LapRLSR modeling algorithm was applied to the online monitoring of the concentration of salvia acid B in the column separation procedure of TCM manufacturing, and the results demonstrated that its prediction capability outperformed PLS and Regularized Least Square Regression method.