Peramalan Penerimaan Pajak Daerah Menggunakan ARIMA, LSTM, dan XGBoost: Studi Kasus Kabupaten Bantul
DOI:
https://doi.org/10.30998/sosioekons.v18i2.4200Keywords:
Pendapatan asli daerah, Peramalan, ARIMA, XGBoost, tapping boxAbstract
Pendapatan Asli Daerah (PAD) Kabupaten Bantul yang bersumber dari pajak hotel, restoran, dan parkir telah dipungut secara elektronik melalui perangkat tapping box. Namun, data transaksi yang telah terkumpul selama bertahun-tahun belum dimanfaatkan secara optimal untuk mendukung peramalan penerimaan daerah. Hasil wawancara dengan Badan Pengelolaan Keuangan, Pendapatan, dan Aset Daerah (BPKPAD) Kabupaten Bantul menunjukkan bahwa proses estimasi penerimaan selama ini masih dilakukan berdasarkan intuisi dan analisis tren sederhana menggunakan aplikasi spreadsheet. Penelitian ini bertujuan untuk mengidentifikasi model peramalan yang paling akurat untuk penerimaan pajak hotel, restoran, dan parkir dengan membandingkan metode ARIMA dan SARIMA sebagai pendekatan statistik, Long Short-Term Memory (LSTM) sebagai pendekatan deep learning, serta XGBoost sebagai pendekatan machine learning berdasarkan kerangka kerja CRISP-DM. Data yang digunakan terdiri atas sekitar 4,8 juta transaksi tapping box periode 2020–2025 yang diolah menjadi data deret waktu dan dievaluasi menggunakan metode rolling-origin backtest dengan indikator Root Mean Square Error (RMSE) dan Mean Absolute Percentage Error (MAPE). Hasil penelitian menunjukkan bahwa tidak terdapat satu model yang secara konsisten memberikan akurasi terbaik pada seluruh sektor dan horizon peramalan. XGBoost menghasilkan akurasi tertinggi pada sektor restoran dengan nilai MAPE terendah sebesar 1,86%, sedangkan ARIMA(1,1,1) menjadi model terbaik pada sektor parkir dengan nilai MAPE berkisar antara 7% hingga 9%. Pada sektor hotel yang memiliki tingkat volatilitas tinggi, model Seasonal Naïve memberikan hasil terbaik dengan nilai MAPE antara 10% hingga 14%. Temuan ini menunjukkan bahwa pemilihan model peramalan harus disesuaikan dengan karakteristik masing-masing sektor serta menegaskan pentingnya kualitas data tapping box dalam mendukung perencanaan penerimaan daerah yang lebih akurat dan berbasis data.
Downloads
References
Abdullah Sani, H. A., & Jaafar, N. I. (2026). Exploring the impact of IoT on governance and public service transformation: evidence from Malaysia’s public sector. Smart and Sustainable Built Environment, 15(5), 1991–2014. https://doi.org/10.1108/SASBE-10-2024-0453
Alsharef, A., Aggarwal, K., Sonia, Kumar, M., & Mishra, A. (2022). Review of ML and AutoML Solutions to Forecast Time-Series Data. Archives of Computational Methods in Engineering, 29(7), 5297–5311. https://doi.org/10.1007/s11831-022-09765-0
Chen, L., & He, R. (2024). Does digital tax enforcement drive corporate digitalization? Evidence from the Golden Tax Project III in China: A pre-registered report. Pacific-Basin Finance Journal, 83, 102242. https://doi.org/10.1016/j.pacfin.2023.102242
David, A., Yigitcanlar, T., Li, R. Y. M., Corchado, J. M., Cheong, P. H., Mossberger, K., & Mehmood, R. (2023). Understanding Local Government Digital Technology Adoption Strategies: A PRISMA Review. Sustainability, 15(12), 9645. https://doi.org/10.3390/su15129645
He, Y., & Yi, Y. (2023). Digitalization of tax administration and corporate performance: Evidence from China. International Review of Financial Analysis, 90, 102859. https://doi.org/10.1016/j.irfa.2023.102859
Jothi, L. S. (2025). Time Series Analysis Using LSTM Networks and Its Application to Financial Forecasting (pp. 19–58). https://doi.org/10.4018/979-8-3373-1200-2.ch002
Kim, N., & Schwartz, Z. (2013). The accuracy of tourism forecasting and data characteristics: a meta-analytical approach. Journal of Hospitality Marketing & Management, 22(4), 349–374.
Kumbasaroglu, H. (2026). Global Tea Production Forecasting Using ARIMA Models: A Multi-Country Time-Series Analysis (1961–2028). Sustainability, 18(2), 1005. https://doi.org/10.3390/su18021005
Lee, S. (2025). Digitalization and Domestic Revenue Mobilization: Governmental Responsiveness and Capabilities Matter. Governance, 38(4). https://doi.org/10.1111/gove.70076
Li, W., & Li, D. (2024). How does technological innovation affect the fiscal decentralization of local governments? Evidence from China. Heliyon, 10(9), e30132. https://doi.org/10.1016/j.heliyon.2024.e30132
Li, X., Li, K., Shen, S., & Tian, Y. (2023). Exploring Time Series Models for Wind Speed Forecasting: A Comparative Analysis. Energies, 16(23), 7785. https://doi.org/10.3390/en16237785
Li, Y., Wu, K., & Liu, J. (2023). Self-paced ARIMA for robust time series prediction. Knowledge-Based Systems, 269, 110489. https://doi.org/10.1016/j.knosys.2023.110489
Lubis, M., Saputra, M., & Agustika Nurtrisha, W. (2021). Financial Technology Development Framework for Prosperity of the Nation and Potential Direction. Proceedings of the 9th International Conference on Computer and Communications Management, 212–218. https://doi.org/10.1145/3479162.3479194
Mahmud, Windu Gata, Hafifah Bella Novitasari, Sigit Kurniawan, & Dedi Dwi Saputra. (2024). Indonesian Government Revenue Prediction Using Long Short-Term Memory. Inspiration: Jurnal Teknologi Informasi Dan Komunikasi, 14(1), 111–124. https://doi.org/10.35585/inspir.v14i1.67
Ongo Nkoa, B. E., & Song, J. S. (2022). Les canaux de transmission des effets des TIC sur la mobilisation des recettes fiscales en Afrique. African Development Review, 34(S1). https://doi.org/10.1111/1467-8268.12650
Rich, R., Bram, J., Haughwout, A., Orr, J., Rosen, R., & Sela, R. (2005). Using Regional Economic Indexes to Forecast Tax Bases: Evidence from New York. Review of Economics and Statistics, 87(4), 627–634. https://doi.org/10.1162/003465305775098215
Scherer, M. (2025). Forecasting Public Expenditure Using LSTMs: A Multi-Task Learning Approach for Volatile Data. 2025 30th International Conference on Automation and Computing (ICAC), 1–6. https://doi.org/10.1109/ICAC65379.2025.11196732
Song, X., Deng, L., Wang, H., Zhang, Y., He, Y., & Cao, W. (2024). Deep learning-based time series forecasting. Artificial Intelligence Review, 58(1), 23. https://doi.org/10.1007/s10462-024-10989-8
Thayyib, P. V., Thorakkattle, M. N., Usmani, F., Yahya, A. T., & Farhan, N. H. S. (2023). Forecasting Indian Goods and Services Tax revenue using TBATS, ETS, Neural Networks, and hybrid time series models. Cogent Economics & Finance, 11(2). https://doi.org/10.1080/23322039.2023.2285649
Valle-Cruz, D., & García-Contreras, R. (2025). Towards AI-driven transformation and smart data management: Emerging technological change in the public sector value chain. Public Policy and Administration, 40(2), 254–275. https://doi.org/10.1177/09520767231188401
Wang, X., & Su, C. (2025). Deep learning for cancer detection based on genomic and imaging data: A comprehensive review. Cancer Management and Research, 2089–2125.
Williams, D. W., & Kavanagh, S. C. (2016). Local government revenue forecasting methods: Competition and comparison. Journal of Public Budgeting, Accounting & Financial Management, 28(4), 488–526.
Yu, X. (2024). Construction of Local Fiscal Revenue Forecasting Model Based on SVR Model. Applied Mathematics and Nonlinear Sciences, 9(1). https://doi.org/10.2478/amns-2024-0784
Yukhno, A. (2024). Digital Transformation: Exploring big data Governance in Public Administration. Public Organization Review, 24(1), 335–349. https://doi.org/10.1007/s11115-022-00694-x
Zhao, Y., Wang, Y., Liu, J., Xia, H., Xu, Z., Hong, Q., Zhou, Z., & Petzold, L. (2021). Empirical quantitative analysis of COVID-19 forecasting models. 2021 International Conference on Data Mining Workshops (ICDMW), 517–526.
Zhu, Y., Wang, Y., Liu, T., & Sui, Q. (2018). Assessing macroeconomic recovery after a natural hazard based on ARIMA—a case study of the 2008 Wenchuan earthquake in China. Natural Hazards, 91(3), 1025–1038. https://doi.org/10.1007/s11069-017-3163-1
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Brian Rizadhani Latuconsina, Prof. Dr. Achmad Nizar Hidayanto, S.Kom., M.Kom. (Author)
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
