Answers are grounded in this paper and cite their sources.
Classification model identifies children at risk of school dropouts, suggesting a need for preventive measures.
The phenomenon of school dropouts among adolescents is one of the most serious challenges in the education sector in Indonesia, particularly in West Java Province, which had the highest rate of school dropouts in Indonesia in 2024. To identify children at risk of dropping out of school, this study developed a classification model using the Multilayer Perceptron (MLP) algorithm. The predictor variables used in this study include age, number of household members, gender, marital status, highest education level of the head of household, residential area, and activity status. This study employs data preprocessing, application of the MLP algorithm using the backpropagation method, and model evaluation using accuracy, precision, recall, and F1-score. The evaluation results of the multilayer perceptron model show an accuracy of 81%, precision in the school dropout class of 28%, recall in the school dropout class of 87%, and an F1-score of 42%. These values indicate that the model is capable of identifying cases of children at risk of dropping out of school, although it is not yet sufficiently accurate in determining who falls into that category. The results of this study are expected to serve as a reference for authorized institutions to undertake preventive measures against school dropout cases using an effective prediction model, as well as a reference for further research in the field of social classification using machine learning. Abstrak. Fenomena putus sekolah pada kalangan remaja adalah salah satu tantangan yang sangat serius dalam sektor pendidikan di Indonesia, khususnya di Provinsi Jawa Barat yang menduduki kasus putus sekolah tertinggi di Indonesia pada tahun 2024. Untuk mengidentifikasi anak-anak yang berisiko putus sekolah, penelitian ini mengembangkan model klasifikasi menggunakan algoritma Multilayer Perceptron (MLP). Variabel prediktor yang digunakan pada penelitian ini meliputi usia, jumlah anggota rumah tangga, jenis kelamin, status pernikahan, pendidikan tertinggi kepala keluarga wilayah tempat tinggal, dan status kegiatan. Penelitian ini menggunakan tahapan data preprocessing, penerapan algoritma MLP dengan metode backpropagation, serta evaluasi model menggunakan accuracy, precission, recall, dan f1-score. Hasil evaluasi pada model multilayer perceptron yang dibuat menunjukkan Accuracy sebesar 81%, Precission pada kelas putus sekolah sebesar 28%, Recall pada kelas putus sekolah sebesar 87%, dan F1-Score sebesar 42%. Nilai-nilai tersebut menunjukkan bahwa model mampu mengenali kasus anak yang berisiko putus sekolah, meskipun belum cukup tepat dalam menentukan siapa yang termasuk dalam kategori tersebut. Hasil penelitian ini diharapkan menjadi referensi bagi lembaga yang berwenang untuk melakukan upaya pencegahan terjadinya kasus putus sekolah dengan model prediksi yang efektif serta menjadi referensi untuk pengembangan penelitian selanjutnya dalam bidang klasifikasi sosial menggunakan machine learning.
No takes yet. Share an insight, caveat, or question.
Ramdhani et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: