Pemilihan Neuron LSTM dan LSTM Bayesian Optimization Untuk Prediksi Curah Hujan Bulanan Berbasis Iklim
DOI:
https://doi.org/10.37859/jf.v15i2.9251
Abstract
Prediksi curah hujan yang akurat penting untuk mendukung ketahanan iklim dan mitigasi risiko hidrometeorologis di wilayah tropis seperti Indonesia. Penelitian ini mengembangkan model prediksi curah hujan bulanan menggunakan Long Short-Term Memory (LSTM) yang dioptimasi dengan Bayesian Optimization (BO). Data yang digunakan meliputi curah hujan bulanan (1983–2024) dari tiga stasiun meteorologi di Pulau Jawa—Tunggul Wulung (Cilacap), Tegal, dan Ahmad Yani (Semarang)—serta variabel prediktor global seperti SOI, Nino 3.4, IOD, WNPMI, AUSMI, dan SST. Model LSTM terdiri dari empat lapisan bertingkat dengan Dropout dan BatchNormalization untuk mencegah overfitting. BO digunakan untuk menentukan kombinasi hiperparameter optimal pada setiap lapisan jaringan. Evaluasi menggunakan Root Mean Squared Error (RMSE) dan Mean Absolute Error (MAE) menunjukkan bahwa LSTM-BO memberikan peningkatan kinerja dibanding LSTM standar di ketiga lokasi. Di Cilacap, RMSE menurun dari 160,51 mm menjadi 148,87 mm, dan MAE dari 126,77 mm menjadi 115,14 mm. Di Tegal, RMSE turun dari 89,00 mm menjadi 86,19 mm, dan MAE dari 65,73 mm menjadi 58,84 mm. Di Semarang, RMSE berkurang dari 104,05 mm menjadi 100,21 mm, dan MAE dari 76,32 mm menjadi 70,60 mm. Integrasi time series, deep learning, dan optimasi probabilistik menghasilkan model yang lebih optimal. LSTM cenderung stabil dan konsisten dengan hasil mendekati observasi di sebagian besar bulan, sedangkan LSTM-BO unggul pada bulan atau lokasi tertentu meskipun pada beberapa kondisi prediksinya lebih jauh dari observasi dibandingkan LSTM
Downloads
References
Zhang, G.P., 2003. Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing, 50,pp.159–175.
Liyew, C.M. and Melese, H.A., 2021. Machine learning techniques to predict daily rainfall amount. Journal of Big Data, 8,pp.1–11.
Buslim, N. and Iswara, R.P., 2019. Pengembangan algoritma unsupervised learning technique pada big data analysis di media sosial sebagai media promosi online bagi masyarakat. Jurnal Teknik Informatika, 12(1), pp.79–96.
Lin, Y., Yan, Y., Xu, J., Liao, Y. and Ma, F., 2021. Forecasting stock index price using the CEEMDAN-LSTM model. North American Journal of Economics and Finance, 57, pp.1–14.
Makarova, A., Biedenkapp, G., Lindauer, M. and Hutter, F., 2021. Overfitting in Bayesian optimization: an empirical study and early-stopping solution. NeurIPS Workshop on Neural Architecture Search.
Ulum, D.S.N. and Girsang, A.S., 2022. Hyperparameter optimization of long short term memory using symbiotic organism search for stock prediction. International Journal of Innovative Research and Scientific Studies, 5(2), pp.121–133.
Snoek, J., Larochelle, H. and Adams, R.P., 2012. Practical Bayesian optimization of machine learning algorithms. Advances in Neural Information Processing Systems, 25.
Cabrera, D., Jiménez, F., Vélez, J.F. and Taborda, H., 2023. Rainfall forecasting using a Bayesian framework and long short-term memory multi-model estimation. Earth Science Informatics, 16(2), pp.1373–1388.
Lathika, P. and Sheeba, D.S., 2023. A novel model for rainfall prediction using hybrid stochastic-based Bayesian optimization algorithm. Environmental Science and Pollution Research, 30, pp.92555–92567.
Ardhitama, A. and Sholihah, R., 2013. Model simulasi prakiraan CH bulanan pada wilayah Riau dengan menggunakan input data SOI, SST, NINO 3.4, dan IOD. Jurnal Sains & Teknologi Modifikasi Cuaca, 14(2), pp.95–104.
Tjasyono, B.H.K., 2004. Klimatologi. Bandung: Institut Teknologi Bandung.
NOAA Climate.gov, 2025. Meet ENSO’s neighbor, the Indian Ocean Dipole. [Online] Tersedia di: https://www.climate.gov/ [Accessed 1 May 2025].
Badan Meteorologi, Klimatologi, dan Geofisika (BMKG), 2025. Data Online BMKG. [Online] Tersedia di: https://dataonline.bmkg.go.id/home [Accessed 1 May 2025].
Bureau of Meteorology, 2025. Southern Oscillation Index (SOI). [Online] Tersedia di:https://www.bom.gov.au/climate/enso/soi/ [Accessed 1 May 2025].
NOAA Climate Prediction Center, 2025. ENSO: Cold & warm episodes by season. [Online] Tersedia di:https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php [Accessed 1 May 2025].
NOAA Physical Sciences Laboratory, 2025. NCEP/NCAR reanalysis surface data. [Online] Tersedia di:https://downloads.psl.noaa.gov/Datasets/ncep.reanalysis/Monthlies/surface/ [Accessed 1 May 2025].
NOAA Climate.gov, 2025. Meet ENSO’s neighbor, the Indian Ocean Dipole. [Online] Tersedia di: https://www.climate.gov/[Accessed 1 May 2025].
NOAA Physical Sciences Laboratory, 2025. COBE2 sea surface temperature dataset. [Online] Tersedia di:https://psl.noaa.gov/data/gridded/data.cobe2.html [Accessed 1 May 2025].
Khodarahmi, M. and Maihami, V., 2023. A review on Kalman filter models. Archives of Computational Methods in Engineering, 30, pp.727–747.
Mega, T., Ushio, T., Takahiro, M., Kubota, T., Kachi, M. and Oki, R., 2019. Gauge-adjusted global satellite mapping of precipitation. IEEE Transactions on Geoscience and Remote Sensing, 57(4), pp.1928–1935.
Hyndman, R.J. and Athanasopoulos, G., 2018. Forecasting: Principles and practice. 2nd ed. Melbourne: OTexts.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Muhammad Luthfan Hawali, Walid

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Copyright Notice
An author who publishes in the Jurnal FASILKOM (teknologi inFormASi dan ILmu KOMputer) agrees to the following terms:
- Author retains the copyright and grants the journal the right of first publication of the work simultaneously licensed under the Creative Commons Attribution-ShareAlike 4.0 License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal
- Author is able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book) with the acknowledgement of its initial publication in this journal.
- Author is permitted and encouraged to post his/her work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of the published work (See The Effect of Open Access).
Read more about the Creative Commons Attribution-ShareAlike 4.0 Licence here: https://creativecommons.org/licenses/by-sa/4.0/.










_(1).png)



