Application of visible/near infrared spectroscopy and chemometrics for storage temperature discrimination of bee pollen
In order to investigate a fast and efficient method for the freshness determination of bee pollen, Visible and near infrared (Vis/NIR) spectroscopy combined with least squares-support vector machine (LS-SVM) was applied to determine storage temperature of bee pollen. Five storage temperatures including -18℃, 4℃, 15℃, 25℃ and 40℃ were setted. The Camellia pollens stored for 60 days at different temperature were investigated. Spectra were acquired by an ASD Fieldspec spectrometer, and pretreatments of the optimal smoothing way of moving average with three segments and multiplication scatter correction (MSC) were applied. After principle component analysis of spectra from 400 to 1000 nm, 4 to 20 principal components (PCs) were chosen as the inputs of LS-SVM models, respectively. Results show that the rp 2 of the LS-SVM model with 20 PCs was more than 0.99 in validation set, better than the often-used PLS model. The results indicated that the PCs reflected and represented the main characteristics of bee pollens stored at different temperatures, and the temperature discrimination was successfully implemented using Vis/NIR spectroscopy based on LS-SVM.
Visible/Near Infrared Spectroscopy bee pollen storage temperature principal component analysis least squares support vector machines
JIN Hang-feng HUANG Ling-xia JIN Pei-hua LOU Cheng-fu
College of Animal Sciences,Zhejiang University,Hangzhou 310029,China
国际会议
北京
英文
1-5
2009-10-14（万方平台首次上网日期，不代表论文的发表时间）