Predicting uniaxial compressive strength of oil palm shell concrete using a hybrid artificial intelligence model

The replacement of natural coarse aggregate (NCA) by oil palm shell (OPS) in concrete mixture offers various advantages such as conservation of natural resources, reduction of CO2 emission, and cost saving. However, mechanical properties of OPS concrete are different from natural aggregate concrete (NAC) due to inferior properties of OPS compared with NCA. Therefore, accurate evaluation of uniaxial compressive strength (UCS) of OPS concrete is of vital importance before construction. To this end, a hybrid artificial intelligence (AI) model based on random forest (RF) was proposed to predict UCS of OPS concrete. The hyperparameters of RF were tuned using beetle antennae search (BAS) algorithm modified by Levy flight and self-adaptive inertia weight. The RF model was trained on a dataset collected from published literature. The obtained AI model has high prediction accuracy with correlation coefficient of 0.9588 on the test set. The proposed method is powerful and efficient in prediction of UCS of OPS concrete and therefore paves the way to intelligent construction.