Data‐Driven Water Demand Prediction: Evaluating Artificial Intelligence Models

Water resource management is crucial in contemporary society due to climate change and water scarcity. These challenges are further exacerbated by rapid urbanization, which leads to an increase in water demand in cities caused by factors such as population growth, industrialization, densification, changing consumption patterns, supply network losses, and climate variability. Accurately predicting water demand is vital for sustainable planning. Various techniques have been developed to estimate water demand with varying degrees of accuracy. In this research, three techniques are used: artificial neural networks (ANNs), random forest (RF), and support vector regression (SVR). The results indicate that artificial intelligence methods demonstrate strong potential for enhancing prediction accuracy in the field of water resources management. All models exhibit high accuracy, with ANN showing the best overall performance.