Analyze influence factors in customer’s insurance transaction by decision tree model

In recent years, the development of the digital is rapidly in the world. In a variety of technologies gradually mature, the Internet and mobile device popularization, the IOT and cloud computing services, driving the growth of all kinds of data, so that the data greatly increased and diversified. The value of these data can be used to predict the consumer’s behavior, difference the user groups to study out efficient marketing strategies, and create differentiated competitiveness.In order to predict the consumer’s behavior of buying insurance products, the research collected 4474 insurance transactions from a bank in Taiwan Tainan. After the data pre-processing, the available transaction number is 3430. In these organized transactions, we let the classification of insurance products as the dependent variable, and the attributes of customers as independent variables. Then, using the correlation analysis by chi-squared test to carry out un-relevant factors. Analyzing the influence factors by decision tree machine learning model. According to the analysis result of the decision tree model, the accuracy rate almost close to 70%, and the most important influence factors are the actual insurance fee and currency. These two influence factors can be used as a reference for the bank in Taiwan Tainan to precise the marketing.