ABSTRACT
Elbisy, M.S., 2015. Sea wave parameters prediction by support vector machine using a genetic algorithm.
The prediction of sea wave parameters is important for the planning, design, and operation of coastal and ocean structures. Many empirical methods, numerical models, and soft computing techniques for wave parameter forecasting have been investigated, but such forecasting is still a complex ocean engineering problem. In this study, the support vector machine (SVM) approach, using various kernel functions, is presented for wave parameters prediction in an attempt to suggest a new model with superior explanatory power and stability. A genetic algorithm is used to determine the free parameters of the SVM. The SVM results are compared with the field data and with backpropagation and cascade-correlation neural network models. Among the models, the SVM with a radial basis function kernel provides the best generalization capability and the lowest prediction error and can therefore be most successfully used for the prediction of wave parameters.
