Investigation of Corrosion Inhibition Capability of Pyridazine Compounds via Ensemble Learning

This research focuses on predicting the corrosion inhibition efficiency (CIE) of pyridazine compounds using machine learning techniques based on the quantitative structure–property relationship model. The assessment of various models through performance metrics, including the coefficient of determination (R²), root-mean-square error (RMSE), mean absolute error (MAE), and chi-squared value (χ²), revealed the gradient boosting regressor (GBR) as the most reliable predictor. GBR exhibits excellent performance compared to other ensemble-based models, with an R² value reaching 0.999, indicating an almost perfect fit to the data. Additionally, GBR displays the lowest RMSE (0.18) and MAE (0.14) values, signifying superior prediction accuracy and precision. The very low chi-square value (0.0004) confirms the model's suitability. This investigation also estimated CIE values for four new pyridazine derivative compounds: P1, P2, P3, and P4. Specifically, the GBR model yielded CIE values ranging from 92.81 to 95.80% for the four pyridazine compounds, which closely resembled the experimental CIE determinations. This innovative method makes it easy to estimate the corrosion inhibition capabilities of new compounds even before experimental synthesis. Emphasizing its potential application in engineering, it simultaneously increases its contextual relevance and importance.