Fusion-Based Data-Driven Strength Prediction of Bamboo Fiber-Reinforced Concrete for Sustainable Construction

The study develops a machine learning-based predictive framework to estimate the compressive strength of bamboo fiber reinforced concrete (BFRC). A literature-derived dataset was compiled from key parameters influencing BFRC performance, including bamboo fiber dosage (0.25–1.00%), fine aggregate content (646–649 kg/m³), slump values (44.6–98.6 mm), and curing ages of 7, 14, and 28 days. The dataset indicates average compressive strengths of 30.18 MPa, 38.51 MPa, and 47.30 MPa at the respective curing periods, reflecting progressive hydration and densification of the cementitious matrix. Machine learning models were developed to capture nonlinear relationships between mixture parameters, fiber dosage, and strength development. The results demonstrate strong predictive capability, with model performance showing high accuracy in estimating compressive strength across different curing stages. The analysis reveals that bamboo fiber content significantly influences the mechanical behaviour of BFRC. Strength increases with fiber addition up to a moderate dosage due to effective crack bridging and stress redistribution within the matrix, while excessive fiber content reduces workability and compaction efficiency. The findings illustrate an optimal bamboo fiber dosage of approximately 0.75%, which provides improved compressive strength while maintaining acceptable workability. The predictive framework offers a reliable decision -support tool for optimizing BFRC mix design and promoting sustainable concrete technologies.