Journal: Journal of Intelligent & Fuzzy Systems, vol. where q defines the mean wave overtopping inflow (m 3 /s/m), Rc defines the crest level (m), parameters A and B are best-fit coefficients determined from the experimental data, and Rmax the maximum wave run-up (m).

Keywords: Artificial intelligence, Mamdani fuzzy approach, breakwaters, crest level approximation, 2% wave run-up It is demonstrated that Mamdani Fuzzy Logic model gives smaller error than the regression-based approach. Finally, 13 unique prototype data of Van de Walle are employed to test the proposed model. In order to verify and compare the presented method's applicability, 261 irregular wave run-up data of Van der Meer and Stam are considered for impermeable, permeable and homogenous rubble mound structures. Hence, the goal is to provide an alternative approach that is better than the regression-based approach at estimating 2% wave run-up level. As an alternative, Mamdani Fuzzy Logic model suggested with no restrictive assumptions. Although, in practic e, regression-based approach is widely-used, it is demonstrated herein that a set of restrictive assumptions hinder its use. Economic and safe design of the crest height of these structures require accurate estimation of 2% wave run-up value on breakwater slope. Tarkan Erdik, E-mail: Ībstract: Safe use of densely populated coastal zones depends on the performance of coastal structures against severe wave attack and strong currents.