基于随机森林算法的溶出方法中溶出介质pH预测模型研究OA

The dissolution test plays an important role in the drug research and development,and the prediction of parameters in the dissolution method can help improve the efficiency of the development of dissolution method.In this study,a random forest machine learn-ing algorithm was used to construct a prediction model of the dissolution medium pH in the dissolution method based on 1331 datasets.The hyperparameters were tuned using grid search techniques.The AUC of the optimized random forest classification prediction model was 0.85(Class 1),0.90(Class 2),0.86(Class 3),and 0.87(Class 4),respectively,and the accuracy was 0.76.The pH prediction model of the dissolution medium for the dissolution method constructed by the optimized random forest algorithm can provide a reference for the development of drug dissolution methods and improve the development efficiency.