基于多元统计分的葡萄酒质量评价分析.docVIP

基于多元统计分析的葡萄质量评价分级 张加勇，冯婷婷，王祥玉，巩梦洁，张晓华 （信阳师范学院华锐学院 数学与计算机科学系 河南 信阳 464000） 摘要：本文研究的是葡萄质量评价和分级问题。首先对第一组品酒员和第二组品酒员对27个红葡萄酒样品的打分均值做均值检验和双正态总体方差检验，得出两组打分均值无显著性差异，而方差有显著性差异，且第二组方差较小，因而结果更可信。对28个白葡萄酒样品做类似处理，得到同样结果，从而可知第二组品酒员的打分结果更可信。然后对白葡萄样品的30个理化指标做主成分分析，提取出10个主成分，把它们和第二组品酒员的打分均值放在一起做R型聚类分析，提取出一些代表性指标，根据这些指标再对28个白葡萄样品做Q型聚类分析和主成分综合评价，综合上述结果，得出这些白葡萄分为四级比较合适。对红葡萄样品做类似处理得出这些红葡萄分为五级比较合适。 关键词：均值检验；方差检验；聚类分析；主成分综合评价 中图分类号：O 221．1 文献标识码：A 文章编号：1671—6132 2011 03—0013—05 Based on multivariate statistical analysis of the grape quality evaluation classification ZHANG Jia-yong,FENG Ting-ting ,WANG Xiang-yu, GONG Meng-jie ,ZHANG Xiao-hua Department of mathematics and computer science, huarui college, Xinyang normal university xinyang 464000，china Abstract: This paper studies the grape quality evaluation and classification problem. First, the first group PinJiuYuan and the second group to PinJiuYuan 27 red wine samples of scoring average do mean inspection and double normal population variance test, it is concluded that the mean score two groups had no significant sex difference, and variance have significant difference, and the second formula difference is small, thus the result more reliable. 28 a white wine samples for similar treatment, get the same results, so as to know the second group of PinJiuYuan score results more reliable. Then dialogue grape samples 30 physical and chemical index family composition analysis, the extracted 10 main composition, put them and the second group of PinJiuYuan scoring average together do R type clustering analysis, the extracted some representative index, according to these index again to 28 white grape sample do Q type cluster analysis and principal component comprehensive evaluation, comprehensive the above results, it is concluded that these white grape is divided into four more appropriate. The red grape sample do similar processing concluded that these red grape is divided into a category five more appropriate. Keywords: average examination;