Here is the presentation I gave at the ICMS workshop AI in Risk Assessment and Mitigation at the Bayes Centre in Edinburgh (https://icms.ac.uk/activities/workshop/ai_in_risk_assessment_mitigation/) on March 4, 2026.

I had the opportunity to present some preliminary work on both calibration and fairness of binary classifiers, particularly in the context of insurance pricing algorithms. More specifically, I presented several calibration-based unfairness metrics related to the Sufficiency group fairness criterion and applied two existing mitigation techniques to achieve group-wise calibration fairness on a life insurance dataset predicting mortality scores.