Psychology Analysis and Modeling of User Registration Behavior in English Online Education Application Based on Random Forest Algorithm

Parents have different attitudes towards the application of online education. This paper is based on a real teaching scenario of a famous English online education application. A large number of students are invited to become trial users. After a period of time, users registered as official users are marked as active users, and vice versa, as lost users. This paper collects the characteristic attributes of all users, and uses decision tree and random forest algorithm to build the portrait characteristics and prediction model of lost users. The model has been applied to training data sets and test data sets and achieved good performance. Its profit efficiency was 81.2% and 77.6% respectively.