Implementation of Random Forest Algorithm for Classification of Eligibility For Social Assistance Recipients In Information Systems

This study aims to develop a web-based information system for classifying the eligibility of social assistance (BANSOS) recipients using the Random Forest algorithm in the Bagan Batu Kota Subdistrict. The system is designed to assist local authorities in identifying BANSOS recipients more accurately and efficiently, minimize errors, and enhance distribution fairness. A quantitative research method was employed, with data collection techniques including observation, interviews, literature review, and document analysis. The dataset consists of 1,100 samples with features such as income, family size, and housing conditions. The Random Forest algorithm was implemented by building a classification model based on training and testing data. The evaluation showed a system accuracy rate of 97%, with a classification error of only 3%. The system provides features for recipient data management, field validation, and automated reporting, supporting more precise decision-making. The results of this study are expected to offer a solution for more effective and transparent social assistance distribution.