<p>This study investigated the eccentric inclined load on the foundation in the three-dimensional field to deal with rectangular and square footings, which were not addressed in the two-dimensional field. With the conditions, the cohesive-frictional slope was incorporated with the finite element limit. At the top of that, utilizing machine learning to predict the ultimate bearing capacity of the foundations without using the mechanical theories. The model conducted in this paper is the Rhinopithecus Swarm Optimization Algorithm. The input parameters were explored in this paper, including slope angle, embedment depth ratio, shape ratio, setback ratio, cohesion ratio, internal friction angle, inclined load angle, and eccentric load ratio. The results indicated that the trend of the load-bearing capacity under the effect of the input data is plain. It had a trend suitable for the mechanical theories. Moreover, the failure mechanism was also very suitable; there were no scenarios exhibiting unusual behavior. In addition, the machine learning outcomes reached high accuracy with the train and test data corresponding to R² = 99.99% and 99.95%, respectively, indicating high confidence.</p>