Robust models to predict the secondary compression index of fine-grained soils using multi objective evolutionary polynomial regression analysis

The secondary consolidation settlement occurs in soft-grained soils after the primary consolidation settlement. This settlement affects the serviceability performance of surface and buried structures in the long term, and thus, it is necessary to evaluate it in the design. The secondary consolidation settlement is calculated using the secondary compression index. However, there is a lack of studies on the development of robust models that could be used with confidence to predict the compressibility index without the need to perform consolidation tests. In this paper, original models have been developed to accurately predict the secondary compression index using an advanced data-driven method. The accuracy of the original models has been examined based on a statistical accuracy assessment methodology. It has been found that three of the new models provide accurate prediction, as they scored low overall error and high cumulative frequency at low error. Also, these models scored a high coefficient of determination. Importantly, the new original models outperformed the available simple empirical correlations in the literature. The outcome of the papers is useful to design engineers who conduct settlement assessments of structures built on soft, fine-grained soils.