葡萄酒的评价

根据给出的数据，分析和研究对葡萄酒的评价问题。采用T检验法、逐对比较法和区间估计法判断两组评酒员的差异性和可信度；运用SPSS对酿酒葡萄理化指标进行主成分分析，再结合葡萄酒的评分结果，根据K-均值聚类方法对酿酒葡萄进行分级；利用多元线性逐步回归模型，给出酿酒葡萄与葡萄酒的理化指标之间的联系；通过典型相关分析，得到酿酒葡萄、葡萄酒理化指标对葡萄酒质量的定性影响，同时运行BP神经网络模型，论证能否用葡萄和葡萄酒的理化指标来评价葡萄酒的质量。

Assessment of Wines

This paper analyzed and studied the evaluation of wines according to the given data. Firstly, two group members of liquor difference and credibility are judged by using T test, comparison and interval estima- tion; then principal components of wines were analyzed by the application of SPSS on physicochemical indica- tors of wines, and combined with the wine scores, the wines were graded according to the K- means cluster- ing; thirdly, the connection of physieochemical indexes between vitis vinifera and wines was provided by re- gression model of multiple linear stepwise; Finally, the qualitative effects on the quality of wines were decided by the vitis vinifera and wine physicochemical indexes by the analysis of canonical correlation, meanwhile, the paper also tested whether vitis vinifera and physicochemical indexes could be used to evaluate the quality of wines by running neural network model BP.