Prediction of Credit Risk in Electronic Commerce Financial Industry Based On Decision Tree Method

Abstract

Decision tree method is one of the most popular data mining technologies, which is mainly used for classification and prediction. Decision tree method is a sample-based inductive classification and decision-making method. It has many advantages, such as small amount of calculation, fast speed, high classification accuracy and easy understanding of classification rules. With the rise of various Internet consumer loans, customer credit overdue risk prediction has become an important research direction in the e-commerce finance industry. Under the background of big data, it is of practical significance for credit loan decision-making to establish a comprehensive and unified data platform to integrate multi-dimensional data information, to model data in multi-dimension, to predict customer credit risk, and to construct a prediction model of credit overdue risk. This paper takes the customer credit risk of e-commerce financial platform as the research object, and constructs a decision tree to determine the decision-making scheme of customer credit rating in order to maximize the expected return.