Classification of Program Keluarga Harapan Recipient Households in Padang City Using K-Nearest Neighbors

Authors

  • Yurivo Rianda Saputra Universitas Negeri Padang
  • Syafriandi Syafriandi
  • Dony Permana
  • Zilrahmi

DOI:

https://doi.org/10.24036/ujsds/vol2-iss2/167

Keywords:

PKH, KNN, SMOTEENN, RFECV

Abstract

Program Keluarga Harapan (PKH) is a social assistance program from the government aimed at providing social protection in the central government's efforts to promote social welfareas. PKH provides benefits to poor families, especially pregnant women and children, by utilizing various health and education services available. PKH benefits also include people with disabilities and the elderly by maintaining their level of social welfare in accordance with the Constitution and the Nawacita of the Republic of Indonesia. The implementation of PKH that experiences distribution errors needs to be classified to ensure its proper distribution. Classification is performed by comparing the number of  neighbors (k) in K-Nearest Neighbors (KNN). The Synthetic Minority Oversampling Technique Edited Nearest Neighbors (SMOTEENN) is applied to balance classes in the target classification and Recursive Feature Elimination Cross Validation (RFECV) is applied to select attributes in the dataset used. The data source was obtained from SUSENAS 2023 data in Padang City. The research results show that KNN with k = 3 is a good algorithm for classifying households recieiving PKH using 10 attributes. KNN with k = 3 achieves an Accuracy of 91,12%, Precision of 89,29%, and Recall of 96,77%.

Published

2024-05-31

How to Cite

Yurivo Rianda Saputra, Syafriandi Syafriandi, Dony Permana, & Zilrahmi. (2024). Classification of Program Keluarga Harapan Recipient Households in Padang City Using K-Nearest Neighbors . UNP Journal of Statistics and Data Science, 2(2), 187–195. https://doi.org/10.24036/ujsds/vol2-iss2/167

Most read articles by the same author(s)

<< < 1 2 3 4 5 6 7 > >>