Estimation of Beginning Points of Cross-Shore Sandbars Using Artificial Neural Network

Sediment transport is critical for the design of coastal structures. In this paper, beginning points of cross-shore sandbars predicted using artificial neural network (ANN), multi-linear regression (MLR), and Quadratic-Multivariable Regression (Q-MR). The dataset was obtained as a result of a physical model . In experiments, 3 different bed slopes and 5 different grain sizes were used. Bed slope, grain size, wave period, and wave steepness were used as independent variables. The dependent variable was the beginning point of cross-shore sandbars (Xb). Mean Average Error (MAE), Mean Square Error (MSE), Root Mean Square Error (RMSE) results, and correlation and regression results were checked to compare the created models. When the results were compared, it was concluded that the ANN model gave better results than traditional statistical methods.