HYPERPARAMETER MODEL LSTM-GRU UNTUK PREDIKSI PEMETAAN TINGKAT KEBAKARAN HUTAN

Authors

DOI:

https://doi.org/10.31000/jika.v9i1.12882

Abstract

Bencana kebakaran hutan merupakan permasalahan besar bagi pemerintah provinsi Kalimantan Tengah. Langkah eksternal maupun internal telah dilakukan melalui kebijakan publik yang dibuat berupa hasil prediksi atau pemetaan kebakaran hutan dimasa akan datang. Dalam penelitian ini dilakukan pengembangan model untuk prediksi tren dan pemetaan tingkat kebakaran hutan dengan fokus penerapan hyperparameter terhadap kombinasi RNN di dua perangkat dan pengaturan rasio dataset berbeda. Dataset yang digunakan merupakan penggabungan dataset MODIS dan Merra2 sebagai end-to-end multivariate fitur dan target. Penggabungan dataset menggunakan asas interpolasi untuk mendukung kontinuitas kekosongan data. Untuk mencapai tujuan penelitian dilakukan eksperimental sebanyak 12 skenario terhadap 6 set pengaturan hyperparameter dengan evaluasi menggunakan performansi regresi MAE dan RMSE. Temuan penelitian menunjukan model kombinasi LSTM-GRU konsisten memperoleh rata-rata error MAE 2% dan RMSE 6% pada P1 dan P2 dengan nilai performa loss pembelajaran terbaiknya berada pada skenario 7, 10, 11 untuk  pembagian kedua dataset dan skenario 8 di rasio dataset 70:30. Pengujian di perangkat berbeda juga tidak mempengaruhi penurunan error pada model terhadap penerapan hyperparameter kecuali lama runtime pembelajaran model. Hasil penelitian ini memberikan gambaran yang komprehensif terhadap pemilihan parameter terhadap kombinasi model RNN yang ideal berdasarkan pembagian rasio dataset serta memberikan pemahaman tentang penerapan hyperparameter pada perangkat berbeda.

References

Chen, Z., Zhang, C., Li, W., Gao, L., Liu, L., Fang, L., & Zhang, C. (2023). Fire danger forecasting using machine learning-based models and meteorological observation: a case study in Northeastern China. Multimedia Tools and Applications, 83(22), 61861–61881. https://doi.org/10.1007/s11042-023-15881-1

Dhake, H., Kashyap, Y., & Kosmopoulos, P. (2023). Algorithms for Hyperparameter Tuning of LSTMs for Time Series Forecasting. Remote Sensing, 15(8). https://doi.org/10.3390/rs15082076

Dzulhijjah, D. A., Majid, M. N., Alwanda, A. Y., Kusuma, D. C., Zakaria, F., Kusrini, K., & Kusnawi, K. (2023). Comparative Analysis of Hybrid Long Short-Term Memory Models for Fire Danger Index Forecasting with Weather Data. 2023 6th International Conference on Information and Communications Technology (ICOIACT), 165–170. https://doi.org/10.1109/ICOIACT59844.2023.10455879

FIRMS Fire Information for Resource Management System. (n.d.). MODIS Active Fire Data. Https://Firms.Modaps.Eosdis.Nasa.Gov/Country/.

GES DISC. (n.d.). Merra2 M2TMNXFLX. Https://Goldsmr4.Gesdisc.Eosdis.Nasa.Gov/Data/MERRA2_MONTHLY/M2TMNXFLX.5.12.4/.

Hema Yurismi, A., Daniah, R., & Wicaksa, A. (2022). Upaya Pemerintah Indonesia Dalam Mengatasi Kebakaran Hutan dan Lahan di Pulang Pisau, Kalimantan Tengah. Interdependence Journal of International Studies, 3(2), 82–96. https://doi.org/10.54144/ijis.v3i2.57

Jamshed, M. A., Theodorou, C., Kalsoom, T., Anjum, N., Abbasi, Q. H., & Ur-Rehman, M. (2022). Intelligent computing based forecasting of deforestation using fire alerts: A deep learning approach. Physical Communication, 55. https://doi.org/10.1016/j.phycom.2022.101941

Kadir, E. A., Kung, H. T., AlMansour, A. A., Irie, H., Rosa, S. L., & Fauzi, S. S. M. (2023). Wildfire Hotspots Forecasting and Mapping for Environmental Monitoring Based on the Long Short-Term Memory Networks Deep Learning Algorithm. Environments, 10(7), 124. https://doi.org/10.3390/environments10070124

Listia Rosa, S., Abdul Kadir, E., Syukur, A., Irie, H., Wandri, R., & Fikri Evizal, M. (2022). Fire Hotspots Mapping and Forecasting in Indonesia Using Deep Learning Algorithm. 2022 3rd International Conference on Electrical Engineering and Informatics (ICon EEI), 190–194. https://doi.org/10.1109/IConEEI55709.2022.9972281

Mao, Y., Pranolo, A., Wibawa, A. P., Putra Utama, A. B., Dwiyanto, F. A., & Saifullah, S. (2022). Selection of Precise Long Short Term Memory (LSTM) Hyperparameters based on Particle Swarm Optimization. Proceedings - International Conference on Applied Artificial Intelligence and Computing, ICAAIC 2022. https://doi.org/10.1109/ICAAIC53929.2022.9792708

PPID KLHK. (2023, December 31). Kinerja Pengendalian Kebakaran Hutan dan Lahan Tahun 2023. Https://Ppid.Menlhk.Go.Id/.

Rini Rahayu. (2023, October 6). 3.436 Titik Karhutla Periode Januari-Oktober, 9.769 Hektare Lahan Terbakar. Https://Www.Detik.Com/Sulsel/Berita/d-6969318/3-436-Titik-Karhutla-Periode-Januari-Oktober-9-769-Hektare-Lahan-Terbakar.

Samadi Miandoab, P., Saramad, S., & Setayeshi, S. (2023). Respiratory motion prediction based on deep artificial neural networks in CyberKnife system: A comparative study. Journal of Applied Clinical Medical Physics, 24(3). https://doi.org/10.1002/acm2.13854

SINDO News. (2023, October 12). Kebakaran Hutan dan Lahan Menjadi Bencana Tahunan di Indonesia, Apa Solusianya? Https://Www.Youtube.Com/Watch?V=j7C4z9vASq4&t=164s.

Zeng, C., Ma, C., Wang, K., & Cui, Z. (2022). Parking Occupancy Prediction Method Based on Multi Factors and Stacked GRU-LSTM. IEEE Access, 10, 47361–47370. https://doi.org/10.1109/ACCESS.2022.3171330

Zhang, H., Zheng, Z., & Wen, G. (2022). Wildfire Monitring Based on LSTM and Deep Learning. IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium, 2682–2685. https://doi.org/10.1109/IGARSS46834.2022.9883872

Zhang, W., Lim, M. K., Yang, M., Li, X., & Ni, D. (2023). Using deep learning to interpolate the missing data in time-series for credit risks along supply chain. Industrial Management and Data Systems, 123(5). https://doi.org/10.1108/IMDS-08-2022-0468

Downloads

Published

2025-01-31

Issue

Vol. 9 No. 1 (2025): JIKA (Jurnal Informatika)

Section

Articles

License

License and Copyright Agreement

In submitting the manuscript to the journal, the authors certify that:

  • They are authorized by their co-authors to enter into these arrangements.
  • That it is not under consideration for publication elsewhere,
  • That its publication has been approved by all the author(s) and by the responsible authorities – tacitly or explicitly – of the institutes where the work has been carried out.
  • They secure the right to reproduce any material that has already been published or copyrighted elsewhere.
  • They agree to the following license and copyright agreement.

Copyright

Authors who publish with International Journal of Advances in Intelligent Informatics agree to the following terms:

  1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License (CC BY-SA 4.0) that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal. 
  2. Authors are 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 an acknowledgment of its initial publication in this journal.
  3. Authors are permitted and encouraged to post their 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 published work.

Licensing for Data Publication

International Journal of Advances in Intelligent Informatics use a variety of waivers and licenses, that are specifically designed for and appropriate for the treatment of data:

Other data publishing licenses may be allowed as exceptions (subject to approval by the editor on a case-by-case basis) and should be justified with a written statement from the author, which will be published with the article.

Open Data and Software Publishing and Sharing

The journal strives to maximize the replicability of the research published in it. Authors are thus required to share all data, code or protocols underlying the research reported in their articles. Exceptions are permitted but have to be justified in a written public statement accompanying the article.

Datasets and software should be deposited and permanently archived inappropriate, trusted, general, or domain-specific repositories (please consult http://service.re3data.org and/or software repositories such as GitHub, GitLab, Bioinformatics.org, or equivalent). The associated persistent identifiers (e.g. DOI, or others) of the dataset(s) must be included in the data or software resources section of the article. Reference(s) to datasets and software should also be included in the reference list of the article with DOIs (where available). Where no domain-specific data repository exists, authors should deposit their datasets in a general repository such as ZENODO, Dryad, Dataverse, or others.

Small data may also be published as data files or packages supplementary to a research article, however, the authors should prefer in all cases a deposition in data repositories.