Fair, Transparent, and Adaptive Credit Models

Credit decisions balance profitability, fairness, and transparency, and lenders face conditions that change over time. This project applies multi-objective reinforcement learning to credit models that treat each of these goals as an explicit objective. The models adapt as conditions change and keep their decisions transparent. The work received first place in the research work category of the Premio AFIRME-FUNAM 2025 (5th edition).