Geometric Regression and Geometric Mixture Modeling for Analyzing Customer Loan Repayment Frequency

Abstrak

This study aims to analyze the factors influencing the frequency of customer loan payments using geometric regression and geometric mixture models. The data used consists of customer loan data, with the dependent variable being payment frequency and the independent variables comprising demographic factors and loan characteristics. The analysis method used is the Generalized Linear Model (GLM) with a geometric distribution, along with the development of a mixture geometric model to capture data heterogeneity. The results show that the geometric mixture model performs better than the single geometric model, as indicated by a lower AIC value. Factors such as length of employment, number of installments, marital status, premiums, and loan term have a significant influence on loan repayment frequency. This model can serve as a basis for credit risk management decision-making.