浙江农业学报 ›› 2021, Vol. 33 ›› Issue (12): 2320-2329.DOI: 10.3969/j.issn.1004-1524.2021.12.12

• 园艺科学 • 上一篇    下一篇

基于主成分与聚类分析的樱桃番茄品质综合评价

周艳超(), 薛坤, 葛海燕, 陈火英, 刘杨*()   

  1. 上海交通大学 农业与生物学院,上海 200240
  • 收稿日期:2020-09-09 出版日期:2021-12-25 发布日期:2022-01-10
  • 通讯作者: 刘杨
  • 作者简介:* 刘杨,E-mail: liuyangtl@sjtu.edu.cn
    周艳超(1996—),女,河北衡水人,硕士研究生,研究方向为樱桃番茄品质。E-mail: 905451058@qq.com
  • 基金资助:
    上海市农委项目(2019-02-08-00-08-F01106);上海市科委项目(17391900500)

Comprehensive evaluation of cherry tomato quality based on principal component analysis and cluster analysis

ZHOU Yanchao(), XUE Kun, GE Haiyan, CHEN Huoying, LIU Yang*()   

  1. College of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai 200240, China
  • Received:2020-09-09 Online:2021-12-25 Published:2022-01-10
  • Contact: LIU Yang

摘要:

为探讨育种材料之间的品质差异,筛选适宜的推广品种,对29份樱桃番茄杂交组合的14个品质相关指标进行测量,运用描述性分析、主成分分析、聚类分析和判别分析对樱桃番茄进行综合评价。结果表明,番茄红素的变异系数最大,达到59%;横径的变异系数最小,仅为11%。主成分分析将14个品质指标简化为5个因子,累积贡献率达79.05%。结合聚类分析,筛选出决定番茄综合品质的6个指标——单果重、糖酸比、硬度、果形指数、VC和番茄红素。综合主成分分析筛选出综合品质得分较高的组合。聚类分析将材料分为6类,结合判别分析构建了樱桃番茄品质判别函数,模型交叉验证显示正确判别率达到79.3%。此研究为樱桃番茄的品质评价提供了新思路。

关键词: 樱桃番茄, 品质评价, 综合主成分分析, 聚类分析, 判别分析

Abstract:

In order to explore the differences in quality between breeding materials and select suitable varieties, this experiment measured 14 quality indicators of 29 cherry tomatoes, using descriptive analysis, principal component analysis, cluster analysis and discriminant analysis to evaluate the quality of cherry tomatoes. The results showed that the coefficient of variation of lycopene was the largest, reaching 59%, and the coefficient of variation of transverse diameter was only 11%. The principal component analysis simplified 14 quality indicators into 5 factors, whose cumulative variance contribution rate reached 79.05%. Combined with cluster analysis, six indicators for determining the comprehensive quality of tomato were selected: single fruit weight, sugar-acid ratio, hardness, fruit shape index, VC and lycopene. Comprehensive principal component analysis was used to screen out the hybrid combination which had the higher comprehensive quality score. Cluster analysis divided the materials into 6 categories. This study had constructed the discriminant function of cherry tomato combined with discriminant analysis. The model cross-validation showed that the correct discriminant rate reached 79.3%. This study provided new ideas for the quality evaluation of cherry tomatoes.

Key words: cherry tomato, quality evaluation, comprehensive principal component analysis (PCA), cluster analysis, discriminant analysis

中图分类号: