Stone columns are widely used columnar methods to improve the bearing capacity of foundation soils and strengthen soft soils. This paper focuses on predicting the bearing capacity of ordinary stone columns using advanced machine-learning techniques. The conventional methods for estimating bearing capacity are often based on simplified assumptions and empirical equations, which may not provide accurate results for complex scenarios. In this study, data for validation are taken from an experimental study reported in the literature. The dataset contains a variety of geotechnical parameters, including cohesion, friction angle, L/D ratio, area replacement ratio, and properties of stone columns. Predictive models are created using state-of-the-art machine learning algorithms, such as random forests, support vector machines, and neural networks. The paper investigated the application of Gradient Boosting Regressor (GBR), Extreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost), and Random Forest for predicting the bearing capacity of ordinary stone columns. By leveraging the power of machine learning, this project aims to contribute to advancing geotechnical engineering practices, ultimately leading to safer and more reliable infrastructure development.