Durability assessment of FRP-to-concrete bonded connections under moisture condition using data-driven machine learning-based approaches

•A comprehensive database is generated to evaluate the durability of FRP-to-concrete connections.•ANN and ensemble methods were able to accurately predict bond strength and failure mode.•A bond strength prediction equation is developed by the M5P model tree algorithm.•A feature importance analysis determined the impact of degrading parameters on the bond.•Duration of moisture condition has the highest influence on the failure mode.