Selecting Best Factors Using Information Gain To Improve Merchant Eligibility Classification

Currently, the financial technology sector growing rapidly. Digital payments significantly drive the digital economy. Payment Service Providers (PSPs) play a crucial role in offering digital payment services and identifying eligible merchants for loan recommendations. The data need further analysis to get accurate load recommendations. This study focuses on supporting the development of merchant businesses, particularly Micro, Small, and Medium Enterprises (MSMEs), by enhancing loan eligibility analysis through the use of machine learning. Information Gain applied to find the best factors to classify merchant eligibility. The factors obtained were then classified using two machine learning algorithms that are Naïve Bayes and K-Nearest Neighbors (KNN). The experiment result shows that classifying merchant eligibility using naïve Bayes and KNN based on selected factors obtained from information gain has better accuracy than based on all factors. These findings demonstrate the importance of selecting relevant variables for more accurate analysis. The results of this study contribute to the understanding of machine learning applications in financial decision-making, offering real solutions for merchant eligibility analysis and supporting the growth of the digital economy