In this study, after examining the geotechnical characteristics, the internal friction angle was estimated based on fines content, relative density, and the corrected standard penetration number using random forest, multilayer perceptron neural network, support vector machine with different kernel functions, and simple and multiple regression, employing data from 13 exploratory boreholes. To evaluate model performance, several indices were used, including the coefficient of determination (R²), variance accounted for (VAF), mean absolute percentage error (MAPE), root mean square error (RMSE), computed performance index (CPI), agreement index (AI), Durbin–Watson index (DI), and the 20% error range index (A20-index). The results of geotechnical investigations revealed that the soil in the study area is predominantly sandy with uniform grading and angular grains, containing a small amount of silt and clay. According to the standard penetration number and existing classifications, the soil is categorized as dense to very dense. The sediments mainly consist of plagioclase, clinopyroxene, epidote, minor quartz, and carbonate fragments. In addition, previous empirical correlations for predicting the internal friction angle were applied and compared with the findings of this research. The results showed that the correlations proposed for sandy soils provide higher applicability in estimating the internal friction angle of the Astaneh-Ashrafieh region. Comparison of the models indicated that the support vector machine with radial basis function kernel outperformed other approaches, achieving a coefficient of determination R² = 0.99, computed performance index CPI = 1.96, RMSE = 0.03, MAPE = 0.06, agreement index AI = 1.00, and A20 = 1.00.