swell-ml-reproducibility - Reproducibility Package for Swell Potential ML Paper

This release contains the full reproducibility pipeline for:

"A Domain-Informed Data Cleaning Framework for Geotechnical Machine Learning: Case Study on Swelling Potential Prediction"

Included: - Raw data reconstruction (PI/C recovery) - DG-DCP domain-informed cleaning - Random Forest models (Model 1 and Model 2) - SHAP and PDP interpretability - Stratified CV and bootstrap uncertainty - All outputs in /outputs/ and plots in /paper_figures/

To reproduce: 1. Clone repo 2. pip install -r requirements.txt 3. Run notebooks in order: 01 → 02 → 03

Published by hssa83 9 months ago