基于主成分聚类分析评价棉花品质的研究

Study on Evaluation of Cotton Quality Based on Principal Component Analysis and Cluster Analysis

采用中国纤维检验局公布的"中国棉花质量分析报告"的8项指标作为原始指标,利用SPSS主成分分析法和聚类分析法对不同地区的细绒棉品质进行分析,筛选出特征根累计贡献率86.931%的3个主成分,对选出的主成分进行系统聚类分析,将棉花品质按相似性分为三类地区。结果表明:各类地区间棉花品质差距较大,第二类地区具有在颜色级、马克隆值指标的优势,第三类地区棉花品质整体较差。研究结果对棉花采购和纺织企业生产具有重要的参考作用。

The paper uses eight indexes of quality analysis report of Chinese cotton released by China Fiber Inspection Bureau as original indexes.SPSS principal component analysis and cluster analysis are applied to analyze the quality of fine staple cotton in different areas and screen 3principal components whose accumulative contribution rate of characteristic root is86.931%.Hierarchical cluster analysis is conducted for selected principal components.Three types of areas are classified according to the similarity of cotton quality.The results show cotton quality in each area has a large gap.The second type of area has such advantages as color level and micronaire value index.Cotton quality in the third type of area is poor.The results has the importamt reference for cotton purchase and production of textile enterprises.