Prediction of Sediment Yields Using a Data-Driven Radial M5 Tree Model

Reliable estimations of sediment yields are very important for investigations river morphology and water resources management. Nowadays, soft computing methods helpful famous regarding the accurate estimation loads. The present study checked applicability radial M5 tree (RM5Tree) model to accurately estimate using daily inputs snow cover fraction, air temperature, evapotranspiration effective rainfall, in addition flow, Gilgit River, Upper Indus Basin (UIB) tributary, Pakistan. results RM5Tree were compared with support vector regression (SVR), artificial neural network (ANN), multivariate adaptive spline (MARS), M5Tree, rating curve (SRC) response surface method (RSM) models. resulting accuracy models was assessed Pearson’s correlation coefficient (R2), root-mean-square error (RMSE) mean absolute percentage (MAPE). prediction during testing period superior ANN, MARS, SVR, RSM SRC R2, RMSE MAPE being 0.72, 0.51 tons/day 11.99%, respectively. predicted suspended peaks better, 84.10% relative accuracy, comparison models, 80.62, 77.86, 81.90, 80.20, 74.58 62.49% accuracies,