Data-Driven Model for Predicting the Compressive Strengths of GFRP-Confined Reinforced Concrete Columns

This paper focuses on the compressive strength of Glass fiber reinforced polymer (GFRP)-confined concrete columns. Data from 114 sets GFRP-confined columns were collected to evaluate researchers’ and proposed model. A data-driven machine learning model was used investigate importance sensitivity parameters affecting strength. The results show that facilitates study confined but suffers a large coefficient variation too high or conservative estimation back propagation (BP) neural network has best accuracy robustness in predicting columns, with only 14.22%, goodness fit for both training testing above 0.9. have an enormous influence are FRP thickness, all parameters, except fracture strain FRP, positively inversely related