PENERAPAN DATA MINING UNTUK MENGKLASIFIKASI PENERIMA BANTUAN PROGRAM KELUARGA HARAPAN MENGGUNAKAN METODE SUPPORT VECTOR MACHINE

Jesika, Amelia
Budianto, Alexius Endy
Nugraha, Danang Aditya

BIMASAKTI
Vol 8 No 1 (2025): BIMASAKTI

The Family Hope Program (PKH) is a governmental initiative in Indonesia designed to decrease poverty and improve the welfare of families. However, the process of identifying eligible families frequently encounters difficulties. To address this, the study applies data mining techniques with the Support Vector Machine (SVM) method to classify prospective PKH recipients in Bangka Leleng Village. The research utilizes 1,039 data samples of recipients from 2019 to 2023, based on five key attributes: age, income, number of dependents, occupation, and home ownership status. Data processing was conducted using Python in the Google Colab environment. The research workflow involved data collection, preprocessing, splitting data for training and testing, analysis, and evaluation using a Confusion Matrix. The test results indicated that the SVM method is highly effective in classifying PKH recipients, achieving an accuracy rate of up to 96%. This optimal accuracy was obtained by employing the RBF kernel, which demonstrated superior performance compared to other kernels. It is anticipated that this research will provide a more efficient and transparent method for determining aid recipients, leading to a more precise distribution of assistance.