Fair & Explainable Credit Scoring

Python — my MSc dissertation project: a credit scoring model for underbanked populations.

Tools & Skills

• scikit-learn
• SHAP
• Fairness metrics
• Model evaluation

Approach

Trained and compared classification models, applied fairness-aware techniques to reduce bias across protected groups, and used SHAP to explain individual predictions.

Outcome

Produced a model that balanced predictive performance with fairness and interpretability, addressing a real gap in credit access.

• ← Back to Python
• Portfolio Home