Erodibility of Nanocomposite-Improved Unsaturated Soil Using Genetic Programming, Artificial Neural Networks, and Evolutionary Polynomial Regression Techniques

Genetic programming (GP) of four levels complexity, including artificial neural networks the hyper-tanh activation function (ANN-Hyper-Tanh), sigmoid (ANN-Sigmoid), evolutionary polynomial regression (optimized with genetic algorithm) (EPR), and intelligent techniques have been used to predict erodibility lateritic soil collected from an erosion site treated hybrid cement. Southeastern Nigeria specifically Abia State is being destroyed by gully erosion, solution which demands continuous laboratory examinations determine parameters needed design sustainable solutions. Furthermore, complicated equipment setups are required achieve reliable results. To overcome constant works needs, prediction becomes necessary. This present research work adopted different metaheuristic soil; classified as A-7-6, weak, unsaturated, highly plastic, high swelling clay content HC utilized in proportions 0.1–12% at rate 0.1%. The results geotechnics aspect shows that HC, a cementitious composite formulated blending nanotextured quarry fines (NQF) hydrated lime activated rice husk ash (HANRHA), improves substantially consistently. outcome models EPR SSE 1.6% R2 0.996 outclassed other techniques, though all showed their robustness ability target (Er) performance accuracy.