Abstract
Coastal zone management, including coastal protection, is one of the main concerns of humanity. Coastal zone management includes protection against extreme waves and floods and affects social and economic life. An important consideration in the construction of coastal structures is the estimation of wave overtopping. In fact, wave overtopping has been responsible for several coastal structure failures. This study sought to evaluate the accuracy of soft computing techniques using Multilayer Perceptron Neural, MLPNN, Network and Random Forest Decision Tree, RFDT, for wave overtopping discharge prediction of vertical coastal structures. The new EurOtop database, which was recently created by gathering a lot of data from various hydraulic testing, was also used to create soft computing techniques for the precise prediction of wave overtopping discharge. Each wave overtopping prediction method's effectiveness has been quantitatively assessed using charts and statistically assessed using accuracy measures. The evaluation's findings demonstrated that the RFDT model provided more accurate predictions than the MLPNN technique.