Urban flooding has become increasingly complex due to rapid urbanization, land-use changes, and extreme rainfall influenced by climate change. This study conducted a Systematic Literature Review (SLR) to analyze the application of Machine Learning (ML) in urban flood prediction and mitigation, particularly in Southeast Asia. A total of 11 empirical articles were selected from the Scopus database using the PRISMA framework, filtered from 217 initial documents based on inclusion criteria covering the period from 2019 to 2026.The review identified three main findings: (1) the implementation of various ML algorithms, such as Random Forest (RF), Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and hybrid models, significantly improved flood prediction accuracy, urban flood vulnerability mapping, and inundation identification; (2) major challenges remain in the availability and quality of hydro-spatial data, spatial bias, model interpretability, and high computational requirements; and (3) the integration of Artificial Intelligence (AI), Geographic Information Systems (GIS), Remote Sensing, Internet of Things (IoT), and cloud computing enhanced real-time flood monitoring systems and supported more effective decision-making.This study concludes that the development of more adaptive, interpretable, and multi-source integrated ML models is essential to support effective, efficient, and sustainable urban flood management in line with the smart city concept.
Copyrights © 2026