PREDIKSI CALON PENERIMA BANTUAN PROGRAM KELUARGA HARAPAN (PKH) MENGGUNAKAN ALGORITMA C4.5 (STUDI KASUS: KECAMATAN BANGKO KABUPATEN ROKAN HILIR)

Abstract

Bangko Sub-district is one of sub-districts in Rokan Hilir Regency that has distributed assistance of Keluarga Harapan (Family Hope) Program since 2007. The program is a conditional social protection program from the central government through distribution of cash assistance to very poor families. In the process of validating prospective recipient data of the program assistance, it is still performed manually which takes a long time. In addition, there are often problems regarding the recipient assistance that do not match criteria for participating in the program. This study aims to build a system that is able to predict potential beneficiaries of the expected family program assistance using the C4.5 algorithm. The attributes used were place of residence, occupation, pregnancy status, dependents of school age children and family members with disabilities. The category of decisions produced was recipients who were predicted to receive and not receive assistance from the program. This system was designed using Unified Modeling Language and developed with the PHP programming language and MySQL DBMS. The results of system testing calculated using confusion matrix against 300 recipient data and not the recipient of the program produced an accuracy rate of 87%.