|
|||||
|
|
| 基于主成分分析和聚类分析的烤烟质量评价模型构建 | |
| 引用本文： | 叶协锋,魏跃伟,杨字熙,陈秀花,黄元炯,李亚娟,李东霞,刘国顺.基于主成分分析和聚类分析的烤烟质量评价模型构建[J].农业系统科学与综合研究,2009,25(3):268-271,276. |
| 作者姓名： | 叶协锋 魏跃伟 杨字熙 陈秀花 黄元炯 李亚娟 李东霞 刘国顺 |
| 作者单位： | 1. 河南农业大学,国家烟草栽培生理生化研究基地,烟草行业栽培重点实验室,河南,郑州450002 2. 河南省烟草公司,河南,郑州,450003 3. 山东平邑烟草公司,山东,平邑,273300 |
| 基金项目： | 国家烟草专卖局重点科研项目 |
| 摘 要： | 为了建立更为科学、客观的烟叶质量评价方法，采用外观质量评价与常规化学成分评价相结合的方法，研究8个外观质量指标和8个化学成分指标，建立了基于主成分分析和聚类分析的Fisher判别函数的烟叶质量评价模型，并对河南省平顶山市42个烟叶样品进行了分类。采用主成分分析方法共提取了6个主成分，并对各个样品6个主成分得分进行聚类分析，同时根据6个主成分得分和聚类分析结果进行判别分析，对比Fisher判别函数和聚类分析的结果，两者有较好的一致性。因此，可以利用这个烟叶评价模型对烟叶进行质量评价。这使得烟叶质量评价方法更加客观，也开辟了烟叶质量评价的新途径。表5，参16。 |
| 关 键 词： | 烤烟 质量评价 主成分分析 聚类分析 判别分析 |
Model of the Quality Evaluation of Flue-cured Tobacco Based on Principal Component Analysis and Cluster Analysis |
|
| YE Xie-feng,WEI Yue-wei,YANG Yu-xi,CHEN Xiu-hua,HUANG Yuan-jiong,LI Ya-juan,LI Dong-xia,LIU Guo-shun.Model of the Quality Evaluation of Flue-cured Tobacco Based on Principal Component Analysis and Cluster Analysis[J].System Sciemces and Comprehensive Studies In Agriculture,2009,25(3):268-271,276. | |
| Authors: | YE Xie-feng WEI Yue-wei YANG Yu-xi CHEN Xiu-hua HUANG Yuan-jiong LI Ya-juan LI Dong-xia LIU Guo-shun |
| Institution: | YE Xie-feng, WEI Yue-wei, YANG Yu-xi, CHEN Xiu-hua, HUANG Yuan-jiong, LI Ya-juan, LI Dong-xia, LIU Guo-shun( 1. National Tobacco Cultivation and Physiology ＆ Biochemistry Research Centre of He＇ nan Agricultural University, Tobacco Cultivation Key Laboratory of Tobacco Monopoly Department, Zhengzhou 450002, China ; 2 He＇nan Tobacco Comparty, Zhertgzhou 450003, China; 3. Pingyi Tobacco Company of Shandong province, Pingyi 273300, China ) |
| Abstract: | To be more scientific and more external,combined with appearance evaluation method and conventional chemical evaluation method,8 appearaace quality characteritics and 8 chemical quality characteristics were researched. And a flue-cured quality evaluation model was es-tablished based on principal component analysis,cluster analysis and Fisher discriminant functions. 42 samples were classified with this model in Ping ding-shan city He' nan Provice. 6 main components were extracted with principal component analysis. 42 samples were clustered by 6 main components scores with cluster analysis. Discriminant analysis used based on 6 main components scores and the result of cluster analysis.The model was acceptable because the results of Fisher discriminant functions and cluster analysis had a well consistent with each other. Ac-cordingly, this study suggested that the model be a new method with more external evaluation for tobacco quality. |
| Keywords: | flue- cured tobacco quality evaluation principal component analysis cluster analysis discriminant analysis |
| 本文献已被 维普 万方数据 等数据库收录！ | |