Hyperparameter Optimization of Boosting Models Using Particle Swarm Optimization for Flood-Prone Area Classification in West Java
DOI:
https://doi.org/10.47701/6yf6xk83Keywords:
flood area classification, ensemble boosting, West Java, particle swarm optimization, remote sensingAbstract
Flooding is the most prevalent natural disaster in Indonesia, causing substantial loss of life and significant economic damage each year. Therefore, an accurate regional flood condition classification system is essential to support early warning efforts. This study integrates daily rainfall data collected from 460 meteorological stations operated by the Meteorology, Climatology, and Geophysics Agency between 2020 and 2025, flood event records obtained from the National Disaster Management Agency, and geospatial features derived from the GLO-30 Digital Elevation Model (DEM) and Sentinel-1 Synthetic Aperture Radar (SAR) imagery as model inputs. Six classification models were systematically developed and evaluated, including baseline XGBoost, LightGBM, and CatBoost models, along with their respective Particle Swarm Optimization (PSO)-optimized variants. The results indicate that the baseline CatBoost model achieved the best performance prior to optimization, whereas PSO-LightGBM outperformed all competing models after optimization, attaining the highest Macro F1-score of 0.5433 and Macro Recall of 0.5350. PSO optimization proved effective in improving the performance of LightGBM and XGBoost but negatively affected CatBoost, whose default configuration was already close to optimal for the dataset used in this study. Furthermore, SHAP analysis confirmed that Runoff Potential, a hydrology domain-informed engineered feature, was the most influential predictor, followed by elevation and SAR backscatter features, highlighting the significant contribution of remote sensing data integration to the model's discriminative capability.
References
Abdel-Fattah, M., Saber, M., Kantoush, S. A., Khalil, M. F., Sumi, T., & Sefelnasr, A. M. (2023). Enhancing flood risk assessment through integration of ensemble learning approaches and physical-based hydrological modeling. Geomatics, Natural Hazards and Risk, 14(1), 2203798. https://doi.org/10.1080/19475705.2023.2203798
Abedi, M., Mohammadzadeh, O., Mehrabani, G. V., Norouzi, H., Ahmadi, A., & Malekian, R. (2022). Flash-flood susceptibility mapping based on XGBoost, random forest and boosted regression trees. Geocarto International, 37(19), 5479–5496. https://doi.org/10.1080/10106049.2021.1920636
Aydin, H. E., & Iban, M. C. (2023). Predicting and analyzing flood susceptibility using boosting-based ensemble machine learning algorithms with SHapley additive explanations. Natural Hazards, 116(3), 2957–2991. https://doi.org/10.1007/s11069-022-05793-y
Badan Nasional Penanggulangan Bencana. (2024). Banjir awal tahun 2024: Resiliensi Indonesia. https://bnpb.go.id
Costache, R., et al. (2025). Flood susceptibility assessment using multi-tier feature selection and ensemble boosting machine learning models. Water, 17(14), 2041. https://doi.org/10.3390/w17142041
Elghouat, A., et al. (2024). Integrated approaches for flash flood susceptibility mapping: Spatial modeling and comparative analysis of statistical and machine learning models. Journal of Water and Climate Change, 15(8), 3624–3646. https://doi.org/10.2166/wcc.2024.726
Kurniadi, A., et al. (2024). Future projections of extreme rainfall events in Indonesia. International Journal of Climatology, 44(4), 1502–1519. https://doi.org/10.1002/joc.8321
Kurugama, A., et al. (2024). A comparative spatial analysis of flood susceptibility mapping using boosting machine learning algorithms in Rathnapura, Sri Lanka. Journal of Flood Risk Management, 17(2), e12980. https://doi.org/10.1111/jfr3.12980
Nearing, G. S., Cohen, D., Dube, V., Gauch, M., Gilon, O., Harrigan, S., Hassidim, A., Klotz, D., Kratzert, F., Metzger, A., Nevo, S., Pappenberger, F., Prudhomme, C., Shalev, G., Shenzis, S., Tekalign, T. Y., Weitzner, D., & Matias, Y. (2024). Global prediction of extreme floods in ungauged watersheds. Nature, 627, 559–563. https://doi.org/10.1038/s41586-024-07145-1
Nikhil, Y., et al. (2025). Identifying dominant river contributions to urban flooding: A scenario-based study of Makassar City. Frontiers in Built Environment, 11. https://doi.org/10.3389/fbuil.2025.1612416
Painter, C. C., et al. (2025). Designing effective flood early warning systems: A review of barriers, best practices, and key characteristics. Journal of Flood Risk Management. https://doi.org/10.1111/jfr3.70145
Priyambodoho, D., et al. (2022). Effects of urban development on regional climate change and flood inundation in Jakarta, Indonesia. Journal of Disaster Research, 17(4), 516–525. https://doi.org/10.20965/jdr.2022.p0516
Rahayu, R., et al. (2025). The influence of flood disasters on economic growth and poverty in Indonesia [Preprint]. ResearchGate. https://www.researchgate.net/publication/388624476
Saber, M., Boulmaiz, T., Guermoui, M., Abdrabo, K. I., Kantoush, S. A., Sumi, T.,
Boutaghane, H., Nohara, D., & Mabrouk, E. (2022). Examining LightGBM and CatBoost models for Wadi flash flood susceptibility prediction. Geocarto International, 37(25), 7462–7487. https://doi.org/10.1080/10106049.2021.1974959
Saha, S., et al. (2024). SAR-driven flood inventory and multi-factor ensemble susceptibility modelling using machine learning frameworks. Geomatics, Natural Hazards and Risk, 15(1), 2409202. https://doi.org/10.1080/19475705.2024.2409202
Saleh, R. A., et al. (2024). A novel voting ensemble model empowered by metaheuristic feature selection for accurate flash flood susceptibility mapping. Geomatics, Natural Hazards and Risk, 15(1). https://doi.org/10.1080/19475705.2024.2360000
Saputra, H., Dasanto, B. D., Tarigan, S. D., Santikayasa, I. P., & Darmawan, Y. (2026). Daily rainfall observations from 460 stations and flood event records across West Java, Indonesia (2020–2025) [Data set]. Mendeley Data. https://doi.org/10.17632/d7j3s64vvr.3
Siddiqui, A. Q., et al. (2025). Flood susceptibility mapping using machine learning and geospatial-Sentinel-1 SAR integration for enhanced early warning systems. Remote Sensing, 17(20), 3471. https://doi.org/10.3390/rs17203471
Szczepanek, R. (2022). Daily streamflow forecasting in mountainous catchment using XGBoost, LightGBM and CatBoost. Hydrology, 9(12), 226. https://doi.org/10.3390/hydrology9120226
Tian, Y., et al. (2019). Towards global remote discharge estimation: Using the few to estimate the many [Preprint]. arXiv. https://arxiv.org/abs/1901.00786
Wang, Y., et al. (2024). Advancing flood susceptibility modeling using stacking ensemble machine learning: A multi-model approach. Journal of Geographical Sciences, 34(8), 1513–1536. https://doi.org/10.1007/s11442-024-2259-2
Xu, Y., et al. (2024). Research on a hybrid model for flood probability prediction based on time convolutional network and particle swarm optimization algorithm. Scientific Reports, 14, 29876. https://doi.org/10.1038/s41598-024-80100-2
Zhang, Q., et al. (2025). Machine learning-based identification of key factors and spatial heterogeneity analysis of urban flooding. Scientific Reports, 15, 23681. https://doi.org/10.1038/s41598-025-08162-4
