by Hongde H
The article I read for this week is about a new big data underwriting models that is introduced by ZestCash. The model helps analyze credit risk in an better accuracy that would allow the company to extend credit to 25 percent and increase repayment from customers by 20 percent.
According to what I read, ZestCash underwrites by combining Google-style machine learning techniques and data analysis, and traditional credit scoring. As a result, the company can offer credit to people who would be mistakenly turned away.
How it works?
First of all, the model starts by targeting thousands of variables / data, then it computes related ones and transform the best into most useful form that will be combined into meta-variable which describes a borrower in a specific aspects; For example, customer behaviors such as fraud, short-term and long-term credit risk, or the amount of money a borrower will likely repay. The meta-variables are later on allocated into different models and will be run through a different method and finally generates factors that contribute to a final decision.
I chose this article because it is related to what we are lectured last week. besides, the models will help reduce credit flaws and help granting loans to people who really need money in a safer and better way.
Source: Rao, Leena (April, 26 2012) “ZestCash Debuts New Big Data Underwriting Models To Determine Consumer Credit Risk”