欢迎访问《中国农学通报》,

中国农学通报 ›› 2014, Vol. 30 ›› Issue (7): 138-141.doi: 10.11924/j.issn.1000-6850.2013-1830

所属专题: 园艺

• 林学 园艺 园林 • 上一篇    下一篇

日照绿茶化学指纹图谱的研究

李庆伟   

  • 收稿日期:2013-07-04 修回日期:2014-01-09 出版日期:2014-03-05 发布日期:2014-03-05
  • 基金资助:
    日照市科技发展计划 “日照绿茶品质提升关键技术研究” (201022)。

Study on Chemical Fingerprint of Rizhao Green Tea

  • Received:2013-07-04 Revised:2014-01-09 Online:2014-03-05 Published:2014-03-05

摘要: 为研究化学指纹图谱和计算机模式识别技术区分日照绿茶产品和全国其他地区绿茶产品的可行性,分别采用主成分分析方法和线性判别分析的方法建立判别模型。主成分分析方法虽不能完全区分,但能够显示它们之间的差异;线性判别方法经数理统计分析有极显著差异,判别模型成立;经验证,建模集的判别正确率为95.7%,验证集的判别正确率为97%。结果表明,采用线性判别分析的方法,能够较好的区分日照绿茶和全国其他地区绿茶,为保护日照绿茶品牌提供依据。

关键词: 调查, 调查

Abstract: Studies the feasibility of using chemical fingerprint and computer pattern recognition to distinguish Rizhao green tea products and other parts of the country. Adopt the methods of principal component and linear discriminant analysis respectively, successfully established discriminant model. Principal component analysis method can't distinguish, but can show the differences between them. Linear discriminant method through mathematical statistics analysis had significant differences, discrimination model was established. Verified, modeling of discriminant accuracy was 95.7%, the discriminant accuracy of validation set was 97%. The result showed that adopt the method of linear discriminant analysis can better distinguish Rizhao green tea and the rest of the country, Provide the basis for protection of Rizhao green tea brand.