Application Of Data Mining In The Analysis Of Family Insurance Purchase Behavior In China

In recent years,with the rise of big data technology,data mining technology has been widely applied to various fields,and has made remarkable achievements in the application,but the application of data mining technology in the insurance business in China is not mature enough,at the same time,China's insurance companies also need to change the traditional marketing methods,improve data utilization value,to explore the digital marketing mode however,the data mining technology can be well applied to the model.In this background,I study the application of data mining algorithms in insurance companies,such as customer identification,and analyzes the purchasing behavior of Chinese insurance products from the perspective of family,and so provide reference for the insurance company for product marketing.Firstly,this paper briefly introduces the data mining theory and the common classification algorithm,and detailed introduces the decision tree and support vector machine model;According to the modeling process of the sample imbalance problem,analyzes the influence of unbalanced samples on classification,introduces the common processing methods,and the paper improves the support vector machine algorithm,using the distance to adjust the misclassification cost of each sample,in order to solve the sample imbalance problem.Then,using the data of Chinese comprehensive social survey in 2013,established various data mining models and improved support vector machine model,and the classification accuracy of various models was compared.The empirical results show that the support vector machine model is a good solution to the problem of unbalanced samples,the prediction of family insurance purchase behavior accurately,can effectively identify customer value;The performance of the improved support vector machine model is better than that of other common data mining models.At the same time,according to the classification rules of decision tree based join misclassification cost model,this paper finds out the main factors influencing the Chinese family whether to purchase insurance products,and puts forward some suggestions for the insurance company for customer identification and product precision marketing.