Data and code for kinetics-guided machine learning modeling of dye adsorption

# Kinetics-Guided Hybrid Machine Learning for Dye Adsorption Modeling This repository contains the raw experimental dataset and reproducible Python pipeline for modeling dye adsorption (MB and MO) under various pH and biochar dose conditions. The workflow covers classical adsorption kinetics fitting, condition-preserving 5-fold cross-validation, pure machine learning (RF, XGBoost), kinetics-guided hybrid machine learning, 8-model benchmarking, and Tree SHAP interpretability analysis. --- ## 1. Directory Structure ```text├── data/│ └── H3PO4_MO_MB_Final.xlsx # Raw experimental dataset├── 1.Kinetic model parameter check.ipynb├── 2.5-Fold Split.ipynb├── 3.GridSearchCV based ML train.ipynb├── 4.Taning and prediction.ipynb├── 5. Hybrid ML setup.ipynb├── 6. 8 model benchmark.ipynb├── 7. SHAP analysis.ipynb├── requirements.txt└── README.md