Author(s)

Tianyu Wu

Corresponding Author

Tianyu Wu

Abstract

This research paper uses medical cosmetic surgery loan as an example, utilizing 1200 people’s data from a typical medical cosmetic surgery loan company from May,2016 to October,2018. It contrasts current data mining model-Decision Tree with traditional qualitative methods (such as whether a person has mortgage before borrowing money, how long a person has lived in one city, and whether the person borrows money from multiple companies at the same time) which companies normally use in determining if a person has higher risk of not paying back money on time and identifies the key factors that can influence a person’s punctuality in giving back money. The main conclusions of my research are that firstly, decision tree is doing a better job in judging whether people are going to pay back their money on time. Secondly, the default rate in big cities is bigger than that in small cities. What’s more, there is threshold existing within the scope of people’s educational level. People have good payback rate beyond that level and have low payback rate below that level.

Keywords

Medical Cosmetic Surgeries, Surgeries Loans, Risk, Decision Tree