Seismic Soil Liquefaction Potential Assessment Using the Multi Gene Entropy Analysis: Fid-Bau Portal

Seismic Soil Liquefaction Potential Assessment Using the Multi Gene Entropy Analysis

Pathak, S. R. / Dalvi, A. N.

Present work illustrates a promising liquefaction evaluation technique by integrating multi gene model in genetic programming (MGGP) with entropy analysis, one of the multi-criteria decision-making (MCDM) tools. The core theme is the dynamic response based approach developed by the authors. MGGP is designed to generate mathematical models of predictor response data that are “multi-gene” in nature wherein the linear coefficients/weights for each of the evolved genes of an individual are estimated using ordinary least squares method. Authors have earlier presented a novel concept of using entropy analysis which determines weights of various parameters responsible for liquefaction which in turn aids in extracting significant soil and seismic parameters. Here in present work instead of carrying out the routine MGGP analysis, an endeavor is made to investigate the feasibility of combining entropy weights with multi gene model to classify liquefied and non-liquefied cases. A multi gene entropy liquefaction model (MGELM) to identify liquefaction susceptibility (LS) is thus formulated. The MGELM performs satisfactorily on 314 case history records. The findings of the present work agree well with earlier research work. The overall accuracy of correct prediction is 81% and thus ensures its credible potential in liquefaction assessment studies.