基于因子分析的马铃薯淀粉质量指标的关联性研究

Correlation Research of Potato Starch Quality Indexes Based on Factor Analysis

为了提高定西市马铃薯淀粉质量，促进淀粉加工企业向品牌化健康发展，研究应用2007～2013年定西市连续7年15家马铃薯淀粉企业供检的105份淀粉质量检测数据，建立SPSS数据录入系统。通过相关分析、因子分析和聚类分析研究了各项质量指标间的关联性，利用马铃薯淀粉各变量间的相关系数矩阵和公因子方差，以K-均值聚类法按提取的3个公因子对企业进行聚类。结果表明，各项质量指标间存在着一定的相关性，部分质量指标间存在着显著的关联性，只需提取3个公因子就足以反映10个指标的信息。因此，可根据最终聚类结果对马铃薯淀粉加工企业进行分类指导，为定西市马铃薯淀粉生产的可持续发展提供决策参考。

Onehundredandfivedataoftenstarchqualityindexesin 15potatostarchenterprisesofDingxiCityintheperiod of 2007 to 2013 were collected and analyzed using SPSS software in order to improve the quality of potato starch and promote the healthy development of starch processing enterprises. The correlation between each quality index was studied through correlation analysis, factor analysis and cluster analysis, and cluster analysis was made using the correlation coefficient matrix and the common factor variance, and K- means clustering method based on the three extracted common factors of enterprises cluster. The results showed that correlation between each quality index existed to some extent, with some correlation significant, and the three extracted common factors was enough to reflect the information of ten indexes. Therefore, some suggestions would be proposed for potato starch processing enterprises according to the final clustering results, and decision-making reference provided for the sustainable development of potato starch production in Dingxi City.