基于多种统计分析模型的葡萄酒评价

以酿酒葡萄和葡萄酒为研究样本，基于二者的相关指标数据，利用统计产品与服务解决方案（SPSS）软件中的主成分分析法 对酿酒葡萄的30个一级理化指标进行筛选，提取8个葡萄酒主要因子作为最终的主要理化指标，达到简化数据降低运算量的目的。运 用Q类层次聚类分析法把酿酒葡萄的27个样品分为3类。运用统计分析系统（SAS）软件结合统计学中典型相关分析方法，以理化指标 和芳香物质为数据基础，对酿酒葡萄与葡萄酒之间的联系进行研究，得到两两之间的线性关系以及指标整体间的线性相关性。

Evaluation of wine based on various statistical analysis models

With wine grape and wine as the research samples, based on the correlation indexes data of the two, the 30 primary physical and chemical indicators in the wine grape were screened and the eight main factors of wine were selected as the main physicochemical indexes by principal compo- nent analysis (PCA) in statistical products and service solutions (SPSS) software, in order to simplify the data and reduce computation. The 27 samples of wine grape were divided into three categories by the Q class hierarchical clustering analysis method. Using statistical analysis system (SAS) software combined with canonical correlation analysis method in statistics, with the physical and chemical indicators and aromatic substances as data base, the correlation between wine grape and grape wine was researched, and the linear relationship between the two and the overall linear correlation of the indicators were obtained.