摘要
随着葡萄酒需求的大量增加,葡萄酒的质量受到了越来越多的关注。一般地,葡萄酒质量采用感官品尝的结果来评定,但是却经常受到多种因素的影响,同时葡萄酒的质量又没有统一的标准,因此葡萄酒的质量评价体系的建立亟待解决,而酿酒葡萄的质量直接决定了葡萄酒的品级。为了得到较好的葡萄酒先要对葡萄进行筛选。基于葡萄的理化指标较多,用灰色关联分析对数据进行初步处理,提取出影响葡萄质量的数个主要理化指标,再运用数据挖掘中的SOM神经网络技术对葡萄进行聚类分析。仿真结果表明:SOM神经网络能够直观准确地将原27类葡萄样品分为7类,且每一类中的葡萄样品均有一定的相似性。
With the increase in the demand for wine,more and more attention has been paid to the quality of wine. In general,the quality of wine is assessed by the results of sensory tasting,but it is often influenced by a variety of factors,and the quality of the wine is not uniform,so the quality of the wine evaluation system needs to be solved,and the quality of wine grapes directly determines the quality of the wine. In order to get a better wine,the grapes should be screened first. Based on the many physical and chemical indexes of grapes,the data are processed by gray correlation analysis firstly,and several main physical and chemical indexes affecting the quality of grape are extracted. And then the SOM neural network technology is used in data mining to cluster analysis of grapes. The simulation results show that the SOM neural network can classify the original 27 grape samples into seven categories intuitively and accurately,and the samples of each grape have some similarity.
出处
《智能计算机与应用》
2017年第6期42-46,共5页
Intelligent Computer and Applications
关键词
灰色关联分析
SOM神经网络
聚类分析
自组织特征映射
gray correlation analysis
SOM neural network
cluster analysis
self-organizing feature mapping
