The Application Of Statistic Methods In Consumer Segments Data Mining

This paper study the application of statistical methods in the consumer segments of data mining technology which based on the company's sales data of bill of lading. Firstly, the author briefly introduces the data mining and customer segmentation related content, Then research in the customer segmentation of data pretreatment and two aspects customer classification. First,I use the principal component analysis and data pretreatment and ordinary pretreatment method for comparison,the ordinary pretreatment method was missing in filling out and noise data, and with principal component analysis to play a good role for the next dimension analysis,and provide relative smaller dataset,bring convenience to analysis; Second,customer classification,compares the decision tree classification and clustering analysis methods,first using decision tree classification, but the effect is not ideal, then,using the cluster analysis method and Means - Ward K,using the method of clustering and Clementine SAS software realization,and two kinds of clustering results are described in detail. The article studies customer segmentation,but from the Angle of statistical methods,This is from two comparison of statistics method used in the segment of the application of data mining,do a little tried.