The Application Of Statistical Methods On The E-commerce Seller Behavior

Nowadays, with the growing popularity of the network, e-commerce, especially C2C online shopping pattern, also ushered in a vigorous development, domestic Taobao, foreign eBay and other online shopping platforms have gradually integrated into and become a part of people’s lives. While online shopping platforms change people’s consumption patterns, they also bring a flood of data, and these data, including customer data, transaction data, product information, web data, which undoubtedly include a great deal of valuable information, it will be very useful for customer management, marketing, risk control, revenue enhancement of shopping platforms if the data is refined, which also makes the statistical techniques come into play. The research will based on the data of an international online shopping platform, using statistical methods to study the seller performance which is an important participant of the online shopping platform, the research includes two sides:One is to use the logistic regression to study the volatility forecast of seller sales ability; The second is to use cluster analysis, which based on principal component analysis, factor analysis, to study the subdivision of the seller. Finally, I combine the results of the two statistical methods to manage seller much more better which is a major innovation of this article.