Predicting clay compressibility using a novel Manta ray foraging optimization-based extreme learning machine model

Abstract

Designing infrastructure founded on very soft deposits requires soil improvement to reduce the compressibility of clay so as to prevent the development of unacceptably high differential settlements. Assessing the likely compression index of clay is critical in determining the extend of the required soil improvement. This research presents extreme learning machine (ELM) models developed using Manta ray foraging optimization (MRFO) for the prediction of the compressibility of clay for soft ground improvement. The results show that the developed MRFO-ELM model predicts the compressibility of clay with less than ± 20 % deviation of the data for 67 % of the specimen and outperforms the prediction accuracy of simple or multiple regression correlations and advanced neural network models currently reported in the literature.