Neural network model to re-rate and analyze consumer credit for Fintechs support vector machine

In recent years, the lowest income population worldwide has considerably increased the demand for credit of low amounts. However, many of the financial entities that provide such amounts do not have granting models that adapt to the specific characteristics of that market. Therefore, the development of granting models that are based on new institutional policies, which integrate quantitative and qualitative information designed exclusively to serve this sector of the population, is relevant. This article presents a methodology for the re-rate of credits from a database corresponding to a financial institution that is dedicated to the placement of resources effectively. For this re-rate a Vector Support Machine with Logistic Kernel was used, which given its flexibility and high classification capacity, allowed generating three granting models, where its results showed the partial relationships that define the granting policies of a given financial entity.