A Review of Machine Learning Techniques for Predicting Rainforest Watersheds and Enhancing Disaster Preparedness

• Published in: 2024 IEEE 1st International Conference on Communication Engineering and Emerging Technologies (ICoCET) pp. 1 - 4
• Main Authors: Pupathy, Sabrina, Azemi, Saidatul Norlyana, Amir, Amiza
• Format: Conference Proceeding
• Language: English
• Published: IEEE 02-09-2024
• Subjects: Accuracy; Artificial Intelligence; Biological system modeling; Computational modeling; Disasters; Floods; Forecasting; Machine Learning; Prediction and Disaster Preparedness; Predictive models; Rainforest Watersheds; Rainforests; Reviews; Watersheds
• DOI: 10.1109/ICoCET63343.2024.10730279

Floods or rainforest watersheds are among the most frequent and destructive natural disasters, causing significant loss of life and economic disruption. Traditional flood modeling approaches often struggle to predict such events accurately, leading to a growing shift towards data-driven methods, particularly Machine Learning (ML) models. These models, leveraging historical climatic data, have emerged as powerful tools for forecasting floods with improved accuracy and cost-efficiency. This paper provides a comprehensive review of the latest advancements in ML-based flood forecasting, highlighting key algorithms and methodologies that have been benchmarked through qualitative analysis for their robustness, accuracy, effectiveness, and speed. The review also delves into major trends in the field, data decomposition, algorithm ensembles, and optimization techniques, which collectively enhance the predictive performance of ML models. By offering insights into the most promising approaches for both long-term and short-term flood or rainforest watershed prediction, this study serves as a valuable resource for hydrologists and climate scientists in selecting and developing effective flood forecasting systems.