Jabir Abubakar Salisu, Hairulnizam Mahdin, Salama A. Mostafa, Muhammad Aamir, Heru Nurwarsito, Ammar Alazab
Floods are the most common and devastating natural disasters, aggravated by climate change and mass urbanization, and pose an urgent need for reliable techniques for prediction and classification to alleviate their situation. Although Machine Learning (ML) and Deep Learning (DL) methods show great potential, unlike existing literature that still mainly focuses on forecasting, prediction or susceptibility mapping, we combine prediction and classification, systematically review datasets and evaluation indicators and establish a model of prediction, method and performance taxonomy by carrying out a systematic literature review on 30 peer-reviewed articles from the period 2020 through 2025, extracted from primary academic databases with explicit search strings and inclusion criteria. The synthesis indicates that the existing ML based ML models, namely Random Forest (RF), Support Vector Machines (SVM) and Artificial Neural Networks (ANN), are functional in structured hydrological data sets. On the other hand, DL models such as Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNN) have a strong performance in extracting spatiotemporal flood dynamics. A broader overview of some important research and methodological gaps — data availability in underrepresented geographic areas, uneven evaluation criteria, lack of Explainable Artificial Intelligence (XAI) implementation, limited interaction of socio-environmental variables — can be gleaned from the review, which highlights potential gaps in the existing literature. We build upon work in this field by mapping, systematically, datasets, methods, and performance metrics in the review. It provides pragmatic proposals for future research and policy-makers to improve flood risk modeling, disaster preparedness, and climate resilience. © 2026, Penerbit UTHM. All rights reserved.
Faculty of Computer Science and Information Technology, Universiti Tun Hussein Onn Malaysia, Parit Raja, Malaysia; Department of Artificial Intelligence Engineering Techniques, College of Technical Engineering, Alnoor University, Nineveh, Mosul, 41012, Iraq; Department of Computer Science, University of Oxford, Oxford, United Kingdom; Faculty of Computer Science, University of Brawijaya, Malang, Indonesia; Centre for Artificial Intelligence Research and Optimisation (AIRO), Torrens University Australia, Adelaide, Australia
