Application of Gradient Boosting combined with metaheuristic algorithms to predict the compressive strength of concrete using manufactured sand

Abstract

This study applies the Gradient Boosting (GB) algorithm combined with the Honey Badger Optimization (HBA) algorithm to predict the compressive strength of concrete using manufactured sand. Utilizing a dataset of 298 experimental samples, the GB-HBA model was developed, and used to analyze factors such as cement, curing age, maximum aggregate size (Dmax), aggregate content, sand fineness modulus, water/binder ratio, water/cement ratio, water content, sand content, and slump. The results indicate that the GB-HBA model accurately predicts compressive strength of concrete using manufactured sand, significantly enhancing the efficiency and durability of the concrete. This research introduces a novel approach for applying machine learning technology in the construction industry, contributing to sustainable development.