Miguel Ferreira

Developed and calibrated Credit Card Admission and Behavioural Scoring Models using Logistic Regression and Decision Trees. Designed and implemented IRB Candidate Mortgage Scoring Models, including RDS construction, data quality validation, and model calibration aligned with regulatory standards. Built forecasting models leveraging unsupervised learning techniques (Time Series clustering and decomposition) to predict credit portfolio behaviour. Conducted ongoing performance monitoring of Scoring and Rating models in production, ensuring compliance, stability, and alignment with internal risk policies. Supported international monitoring of credit models from BCGA and BNU – Macau, maintaining consistent reporting and validation standards.