Detection of Loan Risk Using Fuzzy Association Rules

Abstract

Nowadays, loans put a lender at risk by granting credit to someone with a poor credit score. Furthermore, the number of transactions in the banking sector is continually increasing, and vast data volumes are available that represent consumer behaviour and loan risks. The vast volume of data and information that must be processed and integrated necessitates the use of specific approaches and technologies. As a result, a financial organization's or bank's selection of a new loan application is a critical issue. To predict credit risk patterns, banks have traditionally employed static models incorporating demographic or static characteristics. There are numerous aspects that influence the decision to grant a loan to an application. But employing the fuzzy association rules, Banks and other private financers can be selecting the best applicant with the least amount of risk during the loan process. We invoked the expert knowledge who are specialists in the finance domain to design all fuzzy rules. There are no hard and fast rules to calculate the accurate loan risk. Hence, Fuzzy association rules can provide the best tools to invoke human Knowledge, human perception, vagueness, and imprecision to predict the exact selection result.