Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : JURNAL MEDIA INFORMATIKA BUDIDARMA

Evaluasi Performa Algoritma Naïve Bayes Dalam Mengklasifikasi Penerima Bantuan Pangan Non Tunai Mohammad Mastur Alfitri; Nurahman Nurahman; Minarni Minarni; Depi Rusda
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 7, No 3 (2023): Juli 2023
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v7i3.6151

Abstract

The improvement of the standard of living of the community in Bapinang Hulu Village is carried out through various social assistance programs. However, the realization of the implementation of social assistance programs did not go smoothly. Social jealousy often occurs among the community during the distribution of social assistance. The distribution of assistance is carried out based on the assessment of the village officials and the Village Consultative Body, which is then validated by the heads of RT and RW. The quota provided by the government is often not in accordance with the actual number of eligible recipients in the village. Another difficulty is determining the criteria or attributes used for the selection of Non-Cash Food Assistance recipients. This study aims to obtain a classification model from which the classification pattern can be applied to the population data of Bapinang Hulu Village for the selection of social assistance recipients. To solve this problem, the classification method is applied using the Naive Bayes Algorithm. The research results show that the performance of the Naive Bayes algorithm model before feature selection had the highest accuracy in the 8th test with an accuracy of 89.80%. Meanwhile, after feature selection, the highest accuracy was found in the 3rd test with an accuracy of 88.37%. The feature selection using the Information Gain algorithm reduced the number of attributes from 16 to 6. Therefore, it is known that the highest accuracy is obtained before feature selection, but in selecting social assistance recipients, more criteria need to be applied, which is time-consuming. Meanwhile, after feature selection, only 6 criteria are used to determine social assistance recipients.