DFT-Machine Learning Approach for Accurate Prediction of p.

In this study, we propose a novel method of p prediction in a diverse set of acids, which combines density functional theory (DFT) method with machine learning (ML) methods. First, the DFT method with B3LYP/6-31++G**/SM8 is used to predict p, yielding a mean absolute error of 1.85 p units. Subsequently, such p values predicted from the DFT method are employed as one of 10 molecular descriptors for developing ML models trained on experimental data. Kernel Ridge Regression (KRR), Gaussian Process Regression, and Artificial Neural Network are optimized using three : involving only hyperparameter optimization (HPO), involving HPO followed by a relative contribution analysis (RCA) and recursive feature elimination (RFE), and involving HPO followed by RCA and RFE on an expanded set of composite features. Finally, it is demonstrated that KRR with yields optimal p prediction at an MAE of 0.60 log units. This algorithm was then utilized to predict the p of 37 novel acids. The two most important features were determined to be the number of hydrogen atoms in the molecule and the degree of oxidation of the acid. The predicted p values were documented for future reference.