Code and results for: Explainable machine-learning assessment of DTC-associated vehicle technical condition using real-world OBD-II data (DieselOBD)

Analysis scripts, block boundaries, block-matching distances, per-fold settings, per-fold predicted probabilities, calibration/threshold tables and figures accompanying the manuscript on explainable machine-learning assessment of a DTC-associated technical-condition class from OBD-II data. Source dataset: DieselOBD (https://github.com/AbouAbdallah-Lounis/OBD-Dataset, commit a9c96dc, CC BY 4.0). Vehicle labels V1–V8 are reconstructed blocks and not official identifiers. Library versions: Python 3.13, scikit-learn 1.8.0, XGBoost 3.4.1, SHAP 0.52.0, pandas 3.0.2, NumPy 2.4.4. See README.txt for the run order.

Hali tarjima qilinmagan