<p>Sedimentation modelling is crucial for sustainable watershed management and efficient reservoir operation. This study aims to develop models for sedimentation rates in reservoirs of Madhya Pradesh, India, using four data-driven techniques: linear regression, polynomial regression, support vector machine, and artificial neural networks. The comparative assessment of all models was carried out based on three performance indices: root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). This study reveals that the polynomial regression model indicates promising performance for the available dataset of sedimentation rate modelling in reservoirs of Madhya Pradesh with an RMSE of 0.156, MAE of 0.082, and R² of 0.998, outperforming linear regression (RMSE: 0.957, MAE: 0.688, R²: 0.920), support vector machine (RMSE: 2.238, MAE: 1.168, R²: 0.564), and artificial neural networks (RMSE: 1.039, MAE: 0.766, R²: 0.906). This research presents a practical and interpretable modelling methodology that can enable policymakers and water resource managers to monitor sedimentation and implement sediment control measures.</p>