基于数据挖掘的白酒分类问题研究  

Research on classification of Baijiu based on data mining

作　　者：刘亚清[1] 马艺翔 LIU Yaqing;MA Yixiang(School of Economics and Management,North China University of Technology,Beijing 100144,China)

机构地区：[1]北方工业大学经济管理学院,北京100144

出　　处：《中国酿造》2018年第4期155-158,共4页China Brewing

基　　金：国家自然科学基金项目(71301002)

摘　　要：白酒中含有的成分众多,各成分含量的不同会影响白酒的口感。通过收集整理白酒的11个特征指标数据以及专家打分法提供的评判等级,利用因子分析等数据挖掘方法进行分类预测,也为白酒分类提供了思路。研究发现,模型对于品类6白酒的预测准确度最高(准确度为90%),且白酒中挥发性酸类物质含量的增加有助于提升白酒口感,而酿造用水中氯化物成分会降低白酒口感。Baijiu contains lots of components,which affects the taste of Baijiu,and also provides thoughts for classification of Baijiu.Through collecting 11 characteristic indexes data of Baijiu and the evaluation levels provided by the experts,the Baijiu samples were classified and predicted by data mining methods such as factor analysis method.Results showed that the prediction model had the highest prediction accuracy(90%)for the of grade six.The increase of volatile acids in Baijiu could enhance the taste of Baijiu,while the chloride in brewing water could reduce the taste of Baijiu.

关 键 词：白酒分类 因子分析 数据挖掘 预测 

分 类 号：TS262.3[轻工技术与工程—发酵工程]