K-Nearest Neighbors Approach for Predicting Drug Solubility in Cosolvent Mixtures

This notebook develops a KNN model to predict drug solubility based on its physicochemical properties and system conditions. The process includes data loading and cleaning, hyperparameter optimization of k, and performance evaluation of the final model.

本Jupyter笔记本构建了一款K近邻（K-Nearest Neighbors，KNN）模型，用于基于药物的理化性质与体系条件预测其溶解度。该流程涵盖数据加载与清洗、k值的超参数优化，以及最终模型的性能评估。