P2P Lending Platforms Bankruptcy Prediction Using Fuzzy SVM with Region Information

Abstract

P2P Online lending has enjoyed exponential growth with multifold increases across all main indicators such as the number of customers, market volumes, and business turnovers. However, the P2P lending industry is flawed due to low quality of risk control. In this paper, we focus on Chinese P2P lending platforms and propose a novel method named FSVM-RI, which uses fuzzy SVM algorithm with region information to predict platform bankruptcy. Experiments on real-world datasets show that our proposed method exploits the region information and yields higher classification rate than other state-of-the-art classifiers when outliers and missing values exist in the dataset.