高级检索

    因子、聚类及判别分析在烟叶风格特色评价中的应用

    Application of Factor, Cluster and Discriminant Analysis in the Evaluation of Style Characteristic of Tobacco Leaves

    • 摘要: 为合理评价烟叶风格特征及不同风格特征烟叶区域分类,采用因子、聚类及判别分析相结合的方法,对河南31个产烟县的169个烟叶样品的风格特征指标进行了分析。结果表明,河南烟叶浓香型风格为尚显著;香韵以干草香、焦甜香与焦香为主;河南烟叶的香型与香气状态、香韵中的焦甜香、焦香、烟气浓度呈极显著正相关,与劲头呈显著正相关;采用因子分析方法提取了5个主因子,根据因子总得分进行聚类分析,将31个产烟县分为三类。建立Fisher判别函数评价模型对聚类结果进行判定,判别结果和聚类分析一致性为100.0%。证明这3种方法相结合用于烟叶风格特色分类评价可行性强,效果较好。

       

      Abstract: 169 tobacco samples from 31 tobacco producing counties of Henan were carried out to study tobacco style and characteristics, aiming to provide a scientific basis for the rational evaluation of tobacco-style features and characteristics of different styles of tobacco area classification. Combined with factor analysis method, cluster analysis method and discriminant analysis method. Result: strong flavor type was little, with hay incense, burnt sweetness aroma and burnt aroma as the principal notes; the extreme significant positive correlation of flavor of Henan tobacco with the motion state of aroma, burnt-sweet aroma and burnt aroma of note and smoke consistence, the significant positive correlation with physiological strength; 5 main factors were extracted with factors analysis. 31 tobacco producing counties were divided into 3 categories through cluster analysis based on factors total score. Conclusion: fisher discriminant function evaluation model was established to judge clustering result, and the consistency of judge result and cluster analysis was 100.0%. The above results showed that this method applied in the evaluation of tobacco style features had strong feasibility and good effect.

       

    /

    返回文章
    返回