Abstarct: The application of machine learning in coastal engineering can enhance the understanding of coastal phenomena, improve resource management, and increase the accuracy of analyses. While coastal morphodynamic studies are typically conducted under storm conditions, this research analyzes the coastal area of Narrabeen, Australia, from a different perspective using machine learning techniques. One of the most important aspects of understanding coastal behavior is its ability to recover after storm events, a concept that can be described as equilibrium coastal behavior. Additionally, identifying and characterizing the dominant and influential phenomena affecting coastal equilibrium is highly significant. Hydrodynamic and morphodynamic data were refined using feature selection methods and the removal of low-importance parameters. Various algorithms, including Decision Trees (DT), XGBoost, and Support Vector Machines (SVM), were employed to analyze and identify key parameters influencing three objective functions: shoreline changes, elevation changes of the berm crest, and the horizontal position of the berm crest. According to the results and findings: for shoreline changes, the three most influential parameters were wave power, changes in berm width, and maximum wave height; for elevation changes of the berm crest, the two most important parameters were horizontal changes of the berm crest and wave power; and for the horizontal position of the berm crest, the five key parameters were berm width, berm crest elevation, water level change, berm slope, and wave breaking index. Among the analyzed machine learning algorithms, SVM demonstrated the best performance across all three objective functions, achieving accuracies of 89.02%, 74.36%, and 70.20%, respectively. These findings suggest that the SVM algorithm is reliable for understanding the governing phenomena of coastal features and plays a significant role in assessing morphodynamic changes in the berm, as well as understanding equilibrium behavior, geometric configuration, and initial slope. Furthermore, the results of this study are applicable to sandy beaches with a median grain size (D50) of approximately 0.3 mm.