PREDICTION OF LOAN STATUS USING LOGISTISCS REGRESSION MODEL AND NAÏVE BAYES CLASSIFIER

Abstract

Giving credit to individuals or firm in assumption that they are going to repay the cash with interest may be a business, but this activity has risks. That is called credit risk. Credit risk must be monitored by the lender or creditor to avoid losses that are experienced as a result of the debtor experiencing default. The focus of this research is to construct a model that can determine whether a prospective applicant's credit application is accepted or rejected. The methods used are logistic regression and Naïve Bayes modeling with predictor variables and response variable are categorical data, which is then applied to the case of significant factors in loan status to determine whether the prospective applicant is eligible to be approved or not to be given credit. The dataset consists of gender, married, dependent, education, self-employed, applicant income, co-applicant income, loan amount, loan amount term, credit history, and property area as predictor variables and loan status as response variable. The results of this research show that logistics regression method is better in classifying loan status of applicants in light of the fact that value of accuracy, precision, recall and F1 score of the logistic regression method is greater than Naïve Bayes classifier.