Machine Learning-Based Models for Shear Strength Prediction of UHPFRC Beams

Estimating shear strength is a crucial aspect of beam design. The goal this research to develop calculation technique for ultra-high performance fiber reinforced concrete (UHPFRC) beams. To begin, test database 200 UHPFRC specimens established. Then, random forest (RF) used evaluate the importance influence factors Subsequently, three machine learning (ML)-based models, including artificial neural network (ANN), support vector regression (SVR), and eXtreme-gradient boosting (XGBoost), are proposed compute strength. Results demonstrate that area longitudinal reinforcement has greatest on capacity beams, ten parameters with high (e.g., reinforcement, stirrup strength, cross-section area, span ratio, volume fraction, etc.) selected as input parameters. models ANN, SVR, XGBoost have close accuracy, their R2 0.8825, 0.9016, 0.8839, respectively, which much larger than those existing theoretical models. In addition, average ratios prediction values experimental results 1.08, 1.02, 1.10, respectively; coefficients variation 0.28, 0.21, respectively. SVR model best accuracy reliability. reliability ML-based better calculating