To prepare a credit risk model for increasing the extracting precision of a customer whose credit risk is high.
A credit risk model deciding device 100 decides a credit risk model based on data being the target of analysis sampled from customer data collected from a store terminal 300 through a communication network C. In this case, the credit risk model deciding device 100 executes processing to integrate integer type variables being the explanatory variables of the credit risk model into real numbers, and then adds the explanatory variables by using a GP, and lastly decides the regression coefficients of the respective explanatory variables of the credit risk model. Then, the credit risk model deciding device 100 transmits the acquired credit risk model through a communication network C to a credit risk managing device 200. The credit risk managing device 200 applies risk analytical processing by applying the prepared credit risk model to customer data.
IKUSHIMA TAKAHIRO
KAMIJO FUMIO
