Codes to the publication "Explainable machine learning and feature engineering applied to nanoindentation data" (https://doi.org/10.1016/j.matdes.2025.113897) published in Materials and Design and Dataset "The High-Speed Steel S390 Microclean™ Nanoindentation Dataset" (https://doi.org/10.5281/zenodo.15639081).
If the code helps you, please cite it using its Zenodo version. Please also consider citing the original publication (https://doi.org/10.1016/j.matdes.2025.113897).
The repository is structured as follows:
Explainable_Machine_Learning_Nanoindentation/
│
├── Results/
│ ├── cross-validation/
│ │ ├── *.pkl ➜ Pickled results from the cross-validation workflow
│ │ └── *.ipynb ➜ Jupyter notebooks for plotting and analyzing Cross-Validation results
│ │
│ ├── models/
│ │ ├── *.pkl ➜ Trained Machine Learning models and corresponding SHAP explainers (https://shap.readthedocs.io/en/latest/)
│ │
│ ├── plots/
│ │ └── *.ipynb ➜ Notebooks generating SHAP and other explanatory plots
│
├── Supervised Machine Learning Pipelines/
│ └── *.ipynb ➜ Cross-validation and model training pipelines
│
├── k-means/
│ └── *.ipynb ➜ Clustering analysis using k-means
C.O.W. T. gratefully acknowledges the financial support under the scope of the UFO program (SPM - PN 3022) by the Austrian State of Styria (Land Steiermark - Abteilung 12 Wirtschaft, Tourismus, Wissenschaft und Forschung).
