• EI
  • Scopus
  • 食品科学与工程领域高质量科技期刊分级目录第一方阵T1
  • DOAJ
  • EBSCO
  • 北大核心期刊
  • 中国核心学术期刊RCCSE
  • JST China
  • FSTA
  • 中国精品科技期刊
  • 中国农业核心期刊
  • CA
  • WJCI
  • 中国科技核心期刊CSTPCD
  • 中国生物医学SinoMed
中国精品科技期刊2020
李拥军, 王美玲, 颜鸿飞, 白秀芝, 朱绍华, 戴华. 基于分类回归树CART的湖南黑茶汤色品质的表征研究[J]. 食品工业科技, 2015, (05): 78-81. DOI: 10.13386/j.issn1002-0306.2015.05.007
引用本文: 李拥军, 王美玲, 颜鸿飞, 白秀芝, 朱绍华, 戴华. 基于分类回归树CART的湖南黑茶汤色品质的表征研究[J]. 食品工业科技, 2015, (05): 78-81. DOI: 10.13386/j.issn1002-0306.2015.05.007
LI Yong-jun, WANG Mei-ling, YAN Hong-fei, BAI Xiu-zhi, ZHU Shao-hua, DAI Hua. Study on characterization of dark tea liquor quality based on CART[J]. Science and Technology of Food Industry, 2015, (05): 78-81. DOI: 10.13386/j.issn1002-0306.2015.05.007
Citation: LI Yong-jun, WANG Mei-ling, YAN Hong-fei, BAI Xiu-zhi, ZHU Shao-hua, DAI Hua. Study on characterization of dark tea liquor quality based on CART[J]. Science and Technology of Food Industry, 2015, (05): 78-81. DOI: 10.13386/j.issn1002-0306.2015.05.007

基于分类回归树CART的湖南黑茶汤色品质的表征研究

Study on characterization of dark tea liquor quality based on CART

  • 摘要: 采用色差法分析了红茶、绿茶、乌龙茶及不同产地的黑茶的汤色。根据茶汤色度值筛选出特征变量,建立了湖南黑茶的分类树CART识别模型。结果表明:湖南黑茶与绿茶、乌龙茶和红茶间两两比较,除与乌龙茶Δa*差异不明显外,其他色度值都存在显著差异;对于不同产地的黑茶对比分析,湖南黑茶除与青砖的ΔL*和生普的Δa*差异不大外,与六堡茶、青砖、藏茶、生普和熟普茶的ΔL*、Δa*、Δb*和ΔE*之间有明显差异;采用预测集样本评估模型质量,分类树CART模型对于茶叶类别和产地的识别正确率分别是100%和93.3%。因此,利用茶汤色度值建立起的分类树CART模型,可以得到湖南黑茶很好的分类识别效果。 

     

    Abstract: The chromatic values of Black Tea,Green Tea,Oolong Tea and different origin of Dark tea liquor were analysed.Based on the characteristic variables extracted from the chromaticities,the CART model was established to classify tea area and type. The results were as follows: Comparison difference with green tea,oolong tea and black tea was made.Hunan dark tea and oolong tea in Δa*had no difference,but the other chromaticity values were significantly different among four tpyes of tea. There were significant difference of ΔL*,Δa*,Δb*and ΔE*among Hunan dark tea with Liubao tea,Qing zhuan tea,Tibetan tea,Unzymic Pu- er tea and zymic Pu- er tea,in addition to Qing zhuan tea's ΔL*and Unzymic Pu- er tea's Δa*.The prediction set was used for evaluating the model quality.The accuracy of discriminant functions for type and producing area were respectively reached 100%and 93.3%.Therefore,the better effect of classification and recognition could be obtained by used the CART model based on the chromatic values.

     

/

返回文章
返回