Metaheuristic artificial intelligence (AI): Mechanical properties of electronic waste concrete

Abstract

The appropriate disposal of electronic waste (E-waste) is becoming a serious concern on a global scale. The purpose of present work is to establish a link between the mix design factors and mechanical strength using the metaheuristic based artificial intelligence (AI) Technique known as gene-expression-programming (GEP). The developed dataset includes several input variables i.e., the percentage of e-waste partial substitute, water to cement ratio, specimen age, water absorption and specific gravities of the aggregates, while the compressive strength (CS), flexural strength (FS) and tensile strength (STS) are used as predictive outcome. The established models were assessed using the root mean square error (RMSE), mean absolute error (MAE), objective function, and performance index as well as the regression measure known as the coefficient of correlation (R2). All strength models showed a significant correlation (R2 = 0.94), with the minimum statistical errors (MAE 2.04, RMSE 2.54), (MAE 0.36, RMSE 0.47), and (MAE 0.43, RMSE 0.54) for CS, FS and STS respectively. Furthermore, the parametric and sensitivity analyses were considered for analyzing impact of particular input variables on the performance of outcome. The established machine learning based metaheuristic models can be utilized confidently to use e-waste concrete in a variety of construction purposes.