Claim Missing Document
Check
Articles

Found 31 Documents
Search

Explainable Ensemble Learning for Transparent and Efficient Zakat Scholarship Selection Jasir, Khalid; Safitri, Nadya; Lusa, Sofian; Sensuse, Dana Indra
Eduvest - Journal of Universal Studies Vol. 6 No. 2 (2026): Eduvest - Journal of Universal Studies
Publisher : Green Publisher Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59188/eduvest.v6i2.52639

Abstract

Education financial assistance funded by Zakat, Infaq, and Waqf holds significant potential to support sustainable higher education. Islamic Trust Fund Ar-Raniry faces challenges in the scholarship selection process due to manual processing. This condition leads to operational inefficiency and raises concerns of subjectivity due to a lack of transparency in rejection or acceptance decisions. This study aimed to improve efficiency, objectivity, and transparency of the selection process with a Decision Support System. The study conducted a Decision-Oriented Diagnosis and Feasibility Study to understand the decision-making process. The design of a Decision Support System employed Unified Modeling Language. A prototype created using Low-code/No-code development tools following the Rapid Application Development methodology, while model development adhered to the Cross-Industry Standard Process for Data Mining. The best classification model is a Soft Voting Ensemble of Naïve Bayes, K-Nearest Neighbors, and Support Vector Machine (SVM) with Synthetic Minority Oversampling Technique (SMOTE). The model achieved an Accuracy of 75.63%, a Macro Precision of 72.80%, Minority Class Precision of 65.96%, Macro Recall of 71.16%, Minority Class Recall of 57.41%, F1-score of 71.79%, and AUC ROC of 69.57% in Holdout Testing. Local Interpretable Model-Agnostic Explanation explained the classification results. The main factor affecting classification for the Accepted class is the Fee Level. The implementation resulted in integrated data and automated business processes. Thereby supporting the acceleration of an efficient, objective, and transparent selection process. An implementation strategy was formulated, including parallel conversion, computer-based training, and system support and maintenance.
Co-Authors Abdullah, Puja Putri Aditya, Silfa Kurnia Alfiani, Husna Altino, Iqbal Caraka Anditama, Muhammad Rizky Andriansyah, Chandra Ansis, Ronny Aprilia, Relaci Ariani, Septi Assaf Arief, Assaf Azhari Rangkuti, Herry Baskoro, Aji Brillianto, Bramanti Cahyaningsih, Elin Dana Indra Sensuse Dea Valentina Devit Setiono Editha Putri, Sabrina Elisabeth , Damayanti Elisabeth, Damayanti Erisa Rizkyani Eva Hidayati, Isnina Farhan Fitriya, Ghina Hadjar, Siti Hamid, Ricky Bahari Handoko, Lutvianto Pebri Hudoarma, Fathurahman Ma'ruf Husain Husain Inayah, Suci - Indra Sensuse , Dana Indra Sensuse, Dana Indra, Muhammad Nadhirsyah Indriasari , Sofiyanti Indriasari, Sofianti Iriyanti, Iindra Irmayani, Irni Jallow , Fatoumatta Binta Jannah, Miskol Jasir, Khalid Khusnu Perdani, Mizana Kunto D.A , Himawan Kurnia Aditya, Silfa Masbudi, Handika Mutia Mayanda , Alia Nadya Safitri Nadya Safitri, Nadya Nur Chasanah Nur Chasanah Nurcholis Ramlan Nuswantara, Prasetya Gilang Pradipta, Dhea Junestya Prasetyo Adi Prasetyo Adi Wibowo, Prasetyo Adi Prastowo, Rahardito Dio Pratiwi, Maharani Eka Prima, Pudy Putri, Sabrina Editha Putro , Prasetyo Adi Wibowo Putro, Prasetyo Adi Wibowo Raafigustina, Try Annisa Ramadhan, Yudistira Richi R. Siregar, Jani Ridwan Afandi Rina Rahmawati, Rina Riri Satria Rizki Kurniawan, Rizki Rizky, Fajar Ruliputra, Ricky Nauvaldy Samidi Samidi, Samidi Saputro, Singgih Dwi Sensuse, Dana I. Sihombing , Boy Sandi Kritian Sinaga, Andri Parulian Sitompul, Sabar Sofiyanti Indriasari Sudarto, Reska Nugroho Tereuri, Shabrina Vinc, Richard Wibowo Putro, Prasetyo Adi Wibowo, Wahyu Setyawan Yaziji, Warda