EDA and predictive modeling on a 50,000-row Pakistan road accident dataset.
Analyzes accident patterns, builds a severity prediction model, and uses SHAP for explainability.
- Python, Jupyter Notebook
-
-
Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, SHAP
-
- Clone the repo
-
- Open
analysis.ipynbin Jupyter -
-
Run all cells
-
- 50K row dataset EDA
-
- Accident severity classification
-
- SHAP feature importance
-
- Pre-trained model:
accident_severity_model.pkl
- Pre-trained model:
-
- Open
-