Prediction of Lending Club Loan Defaulters

This article analyzes the loan defaulter’s prediction of Lending Club. Since the emergence of more and more repayment problems brings risks and capital losses to the company, it is crucial for managers to research the relevant factors of loan failure. The data is collected from Kaggle website, and it contains the data from the United States during the period of 2007–2015. In this paper, comparative analysis, group analysis and index analysis are used to analyze the dataset. What’s more, there are three methods to predict the model building, which are Artificial Neural Networks, XG-Boost Classifier and Random Forest Classifier. Additionally, it can be found that the main factors affecting loan defaulters are installment, terms, grades, interest rate and so on. And when establishing the model, it is found that, Artificial Neural Networks (ANNs) algorithm is more suitable for the analysis and prediction of these data. Lending Club's analysis of failed loans can help managers develop strategies to reduce business risks and maximize profits.