Application of support vector machine model in determining the liquefaction trigger of soil under seismic load

Abstract

ABSTRACT

This study presents the results of applying the approach based on artificial intelligence in determining the liquefaction trigger of the soil under an earthquake. In this study, an artificial intelligence model called support machine vector was developed to predict the probability of soil liquefaction. A database of 288 observed soil liquefaction results from the Chi-chi (1999) earthquake was used to train and test the predictive ability of the model. The results of the study are compared with two experimental formulas based on the soil SPT value, showing that the support machine vector model provides superiority in determining the liquefaction trigger of the soil compared with the two methods. The study shows that the support machine vector model is a model capable of predicting very well the possibility of soil liquefaction, and has great potential in solving other problems in the field of construction.

Keywords: Soil liquefaction; SPT value; earthquake; support machine vector.