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

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。

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.