https://www.ejournal.marqchainstitute.or.id/index.php/Jmirte/issue/feed 2026-09-07T18:05:53+00:00 Jonson Manurung Marcha.institute@gmail.com Open Journal Systems <p>Jurnal Manajemen, Informatika, Rekayasa Perangkat Lunak dan Teknologi Komunikasi (Jurnal Mirte)(<strong>e-ISSN:2961-9467</strong>) is a scientific journal and open access journal published by Pusat Penelitian Teknologi, Marqcha Institute, Indonesia. Jurnal MIRTE covers the field of Management, Informatics, Computer Science, Software Engineering, Information Technology and Communication.. The aims of Jurnal MIRTE are to disseminate research results and to improve the productivity of scientific publications. Jurnal MIRTE is published two times a year (Juni and December).</p> https://www.ejournal.marqchainstitute.or.id/index.php/Jmirte/article/view/416 2026-08-26T17:59:13+00:00 Miftahul Fadhlin miftahulfadhlin@gmail.com Suriati Suriati Suriati.ati@gmail.com

<p>The increasing adoption of digital transformation in the public sector has created opportunities to utilize administrative data for predictive decision-making and service planning. The Office of Religious Affairs (Kantor Urusan Agama/KUA) generates historical marriage service data that can potentially be used to estimate future service demand. However, the application of machine learning, particularly Random Forest Regression, to marriage service demand forecasting in religious administrative services remains limited. This study aims to develop and evaluate a Random Forest Regression model for predicting monthly marriage service demand at KUA Bajenis District, Tebing Tinggi City, based on historical service data. A quantitative approach was employed through several stages, including data preprocessing, chronological feature construction, time-based training and testing, model development, prediction, evaluation, and web-based system implementation. The model utilized year, month, previous-month marriage count, holiday conditions, and Hijri month as predictor variables, while the monthly number of marriage service requests was used as the target variable. The Random Forest model was implemented using 100 decision trees with a random state of 42. Experimental results showed that the model achieved a Mean Absolute Error (MAE) of 1.50, Mean Squared Error (MSE) of 3.7401, Root Mean Squared Error (RMSE) of 1.9339, and Mean Absolute Percentage Error (MAPE) of 37.67%. The model predicted an average demand of 6.49 marriage service requests per month for January–June 2025. These findings indicate that Random Forest Regression can serve as a supporting predictive tool for estimating future marriage service demand and assisting KUA in personnel, resource, and service-capacity planning.</p>

2026-09-07T00:00:00+00:00 Copyright (c) 2026 Jurnal Manajemen, Informatika, Rekayasa Perangkat Lunak dan Teknologi Komunikasi