基于MCCV奇异样本筛选和CARS变量选择法对蜂蜜pH值和酸度的近.pdfVIP

182 2011, Vol. 32, No. 08 食品科学 ※分析检测 基于MCCV 奇异样本筛选和CARS 变量选择 法对蜂蜜pH 值和酸度的近红外光谱检测 1 2 , 3 4 4 2 李水芳 ，单 杨 * ，范 伟 ，尹 永 ，周 孜 ，李高阳 ( 1.中南林业科技大学理学院，湖南长沙 4 10004 ；2 .湖南省食品测试分析中心，湖南长沙 4 10025 ； 3.中南大学化学化工学院，湖南长沙 4 10083 ；4 .湖南明园蜂业有限公司，湖南长沙 4 10005) 摘 要：采用Norris 平滑加一阶微分数据预处理，蒙特卡洛交互验证(MCCV) 的奇异样本筛选和CARS(competitive adaptive reweighted sampling)变量选择法，用Kennard-Stone(KS)法划分训练集和预测集，偏最小二乘(PLS) 回归近红 外光谱建模，对蜂蜜pH 值和酸度进行定量分析。pH 值和酸度校正模型的交互验证决定系数(Rcv 2) 、交互验证均方 差(RMSECV) 、预测集决定系数(Rp 2) 、预测均方差(RMSEP)分别为0.85 16 和0.8723 、0.12 14 和2 .1734 、0.8205 和 0 .82 50 、0 . 1196 和2 .4 6 74 。结果表明，该方法适于蜂蜜p H 值的测定，而不宜用于测定蜂蜜酸度。 关键词：近红外光谱；蒙特卡洛交互验证的奇异样本筛选；CA R S 变量选择；蜂蜜；p H 值；酸度 Analysis of pH and Acidity of Honey by Near Infrared Spectroscopy Based on MCCV Outlier Detection and CARSVariable Selection 1 2, 3 4 4 2 LI Shui-fan g ，SHAN Yan g * ，FAN Wei ，YIN Yon g ，ZHOU Zi ，LI Gao-yan g (1. College of Science, Central South University of Forestry and Technology, Changsha 4 10004, China；2. Hunan Food Test and Analy sis Center, Changsha 4 10025, China ；3. School of Chemistry and Chemical Engineering, Central South University, Changsha 4 10083, China ；4. Hunan Mingyuan Honey Products Co.Ltd., Changsha 4 10005, China) Abstract ：The near infrared spectra of honey samples were calculated by the method of Norris smoothing combined with first derivative. The outliers were detected by Monte Carlo cross validation (MCCV), and the variables were selected by competitive adaptive reweighted sampling (CARS). The samples were