<![CDATA[The settlement of bored pile foundations is one of the critical factors affecting the stability of structures, thereby requiring accurate prediction methods. Along with technological advancements, prediction methods have also evolved, creating new opportunities to simplify the estimation process and obtain more precise results. Support Vector Machine (SVM), as one of the machine learning methods, provides an efficient way to predict foundation settlement compared to direct field measurements.This research aims to develop a predictive model for bored pile foundation settlement using Support Vector Machine (SVM) and to validate its accuracy by comparing the predicted settlement results with analytical settlement calculations. The study utilized field test data, namely the Standard Penetration Test (SPT), building load data (Q), and foundation data obtained from various literature sources. The input parameters considered include pile length (L), pile diameter (D), end bearing capacity (Qp), and shaft resistance (Qs), while the output parameter is foundation settlement.The performance of the model was evaluated using the R² and RMSE metrics. The results indicate that, on training data, the model achieved an R value of 0.9943 and an RMSE of 0.1, demonstrating excellent ability in learning data patterns. However, a considerable performance drop was observed on testing data, with an R value of 0.447 and an RMSE of 0.394. This large discrepancy between training and testing performance suggests mild overfitting, where the model performs very accurately on previously seen data but lacks generalization capability when applied to unseen cases]]>
