Modeling of seismic liquefaction data using extreme learning machine

Seismic liquefaction is a major concern for any earthquake prone area. This article employs a new technique based on Extreme Learning Machine (ELM) for determination of liquefaction susceptibility of soil based on Standard Penetration Test (SPT) and Cone Penetration Test (CPT) from the Chi-Chi earthquake. ELM is the modified version of Single-hidden Layer Feed forward Neural Network. Different input combinations have been tried to get best performance. A comparative study has been carried out between the SPT and CPT based models. CPT based models give better performance than the SPT based model. The developed ELM model shows that Cone Resistance (qc) and Peck Ground Acceleration (PGA) are sufficient parameters for determination of liquefaction susceptibility of soil. © 2017 by Nova Science Publishers, Inc.