<![CDATA[One of the strategic programs in Indonesia to tackle poverty is the Family Hope Program (PKH) which is carried out by the government by providing cash to very poor families. The problem that occurs in PKH is the distribution of aid that is still not on target. Therefore this study aims to create a classification model for PKH beneficiaries to overcome these problems. The algorithms used to create a classification model are the Naïve Bayes Classifier (NBC), K-Nearest Neighbor (K-NN), and C4.5. The validation method used is K-Fold Cross Validation (K = 10). The number of attributes used is 33 attributes. The data used to construct the classification model (data after pre-processing) is as much as 378 data on prospective PKH beneficiaries. Based on the experimental results the NBC algorithm produces an accuracy value of 77.51%, the K-NN algorithm (K = 3) produces an accuracy value of 76.72%, the C4.5 algorithm produces an accuracy value of 80.16%. In addition, the C4.5 algorithm succeeded in reducing the number of attributes, from 33 attributes to just 8 attributes, namely: number of household members, fasbab, other houses, gold, fridge, number of rooms, walls, and excreta disposal. This reduces the complexity of the classification model generated by the C4.5 algorithm.]]>
Copyrights © 2023
