Transfer learning enables prediction of steel corrosion in concrete under natural environments

Dataset on carbonation and chloride-induced steel corrosion in cementitious mortars

“…This dataset is valuable for training various ML algorithms to predict steel corrosion parameters, such as concrete resistivity, corrosion potential of steel bars, and corrosion rate of steel bars. Its effectiveness has been demonstrated in previous studies [ 3 , 4 ].…”

“…This dataset includes not only mixture parameters and environmental parameters but also crucial information such as material properties and electrochemical parameters. Based on this dataset, Ji and Ye et al have developed various ML models to predict the corrosion rate of steel in cementitious mortars [ 3 ] and expanded its predictive capability to natural corrosion of reinforced concrete using transfer learning techniques [ 4 ].…”