Linear Genetic Programming, Naïve Bayes algorithm - Their Applications in Geotechnical Engineering

Abstract

Modeling soil and rock pose challenges due to uncertainties in their complex behavior.  In the present study, linear genetic programming and Naïve Bayes are used in classification of liquefied and non-liquefied data. Soil and seismic parameters influencing the soil liquefaction potential are used to develop the models. Genetic Programming is the automatic creation of computer programs to perform a selected task using Darwinian natural selection. Linear genetic programming forms a peculiar subset of genetic programming where computer programs in a population are constituted as successive repetition of instructions from imperative programming language. Naive Bayes methods are supervised learning algorithms by applying Bayes’ theorem with the “naive†assumption of independence among all the sets of the features. Accuracy of results of classification for linear genetic programming, Naïve Bayes were found to be 94.12% and 90.59% respectively.