ENHANCED CREDIT WORTHINESS OF BANK CUSTOMER IN NIGERIA USING MACHINE LEARNING AND DIGITAL NERVOUS SYSTEM

Abstract

This paper presents an enhanced model of machine learning and digital nervous system based on credit worthiness of bank customer in Nigeria. The machine learning have the capability to determine the relevant features and customer’s credit worthiness in efficient manner. The Digital Nervous System (DNS) approach will enable the Credit Risk Management System (CRMS) to capture credit risk information and provide it where it is needed for decision making and when it is needed. The paper aimed at developing an enhanced model for detecting a credit worthiness in a bank. In recent years, customer’s credit worthiness is becoming more crucial for financial organizations. To deal with major problems like noise, data incompleteness and lack inconsistency in loan while building predictive models, this paper proposes a predictive model for detection of credit worthiness. For developing and testing of model a large, real and most recent dataset of credit card, obtained from UCL repository, is proposed. The paper makes it possible for the really availability of more comprehensive credit risk management information and to minimizes decision-making errors with respect to loans. The key focus of the work is on detection of credit worthiness which is defined as the probability of default on the loan or credit from financial organizations like banks. The efficiency of model is demonstrated using confusion matrix on the basis of prediction accuracy and other metrics, against benchmark classifies.