TY-JOURA2-Filponnen-IlaiAU-Zhang-YU-Wang-Yang-YangAU-YANYUUYYUUUYYUUYYYYUUUUYYYUUUUYYYUUUYUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU2UUUUUUUUUUUUUUUUU2UUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU2UUUUUUUUUUUUUUUUUUUUUUUUUUU2UUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUX.AU-Duan、Jin-aoPY-2021DA-2021/01/12TI-FT-NIR光谱SP-8875876VL-2021AB-AngelicaeSincis Radix快速地理源识别和质量评估研究的目的是开发AngelicaeSincis Radix地理分类方法,并使用近红外光谱测定样本中ferlicaAngelicaeSincis Radix使用Fleier变换近红外线光谱分析支持向量机算法用于建立定性模型最优SVM模型对标定集的识别率为100%,对预测集的识别率为83.72%此外,还用FT-NIR制作量化模型预测frulice和Zliustli偏最小回归算法用于建立量化模型协同区间PLS筛选特征光谱段以获取最佳PLSR模型最佳光谱预处理法所建立的最佳PLSR模型确定系数为0.9659和0.9611,而预测确定系数分别为0.9118和0.9206。预测偏差比值两个最终优化PLSR模型大于2结果表明,NIR光谱学结合SVM和PLSR算法可被用于歧视来自不同地理位置的AngelicaeSinensis Radix以获取质量保证和监测。这项研究可用作农业、医药和食品质量评价参考SN-2090-8865UR-https://doi.org/101155/2021/8875876DO-10.1155/2021/8877876JF-化学分析方法杂志PB-HindawiKW-ER