SETTLEMENT PREDICTION FOR THE HIGH FILL EMBANKMENT VIA SUPPORT VECTOR MACHINE

摘要

Support Vector Machine (SVM) could establish nonlinear relation between input data and output data through adaptive learning of measured samples without any assumptions, and the relationship could reflect the potential laws among data more objectively. In this study, SVM was proposed to predict the settlement of high fill embankment, and the Polynomial kernel function and SMO arithmetic were determined. Through the reported instance, the results could suggest the feasibility and serviceability of SVM in geotechnical engineering.