加权最小二乘支持向量机改进算法及其在光谱定量分析中的应用  

Improved Weighted Least Squares Support Vector Machines Algorithm and Its Applications in Spectroscopic Quantitative Analysis

作　　者：吕剑峰[1] 戴连奎[1] 

机构地区：[1]浙江大学工业控制技术国家重点实验室,杭州310027

出　　处：《分析化学》2007年第3期340-344,共5页Chinese Journal of Analytical Chemistry

基　　金：国家高技术研究发展计划(863计划)项目(No.2006AA04Z169);浙江省科技计划项目(No.2005C311042)资助

摘　　要：为克服异常训练样本对校正模型的负面影响,提出了一种加权最小二乘支持向量机(WLS-SVM)的改进算法,解决了原有算法存在的迭代收敛问题,并将其运用于光谱定量分析。实验结果表明与原有算法相比,WLS-SVM改进算法显著增强了对异常样本的检测能力,并大幅度地提高了校正模型的稳健性。An improved algorithm of weighted least squares support vector machines （WLS-SVM） is proposed to overcome the negatine influence of abnornmal traningsamples on calibration models. The improved algorithm solves the iterative convergence problem in the original algorithm and has been used in spectroscopic quantitative analysis. Compared to the original algorithm, the experimental results show that the new algorithm remarkably improves model robustness and the ability to detect abnormal samples.

关 键 词：支持向量机 异常检测 稳健建模 近红外光谱 

分 类 号：O657.3[理学—分析化学]