<![CDATA[Flood events in Indonesia continue to increase in frequency and impact due to high rainfall variability, land-use change, and complex hydrological conditions. Accurate predictive modeling is therefore essential to support flood risk assessment and mitigation planning. This study evaluates the predictive performance of two supervised machine learning algorithms, Support Vector Machine (SVM) and Extreme Gradient Boosting (XGBoost), for flood risk classification. The analysis is conducted using a publicly available dataset comprising 500 samples that represent multiple environmental and spatial factors related to flood occurrence. Data preprocessing includes cleaning, normalization, and feature consistency adjustment prior to model implementation. Both algorithms are trained and tested using the same dataset configuration to ensure objective comparison. Model performance is assessed using accuracy, precision, recall, and F1-score metrics. Experimental results indicate that XGBoost achieves higher accuracy and precision, demonstrating stronger capability in reducing false-positive predictions, while SVM shows relatively higher recall, reflecting better sensitivity in identifying flood-prone cases. Overall, XGBoost provides more reliable predictive performance for flood risk modeling on the dataset used. The findings confirm the effectiveness of machine learning-based approaches for flood risk prediction and highlight the importance of algorithm selection in disaster risk analysis.]]>
Copyrights © 2026
