What's in motion.

Accepted papers, thesis defenses, awards, grants, events, and partnerships. The day-to-day of the lab — at human scale, without the hype.

Archive since 2022 last update

September 2025

02 items
  • 10 sep Defense

    Felipe Bevilaqua defends a master's dissertation on fairness in synthetic data

    In the dissertation "Fairness Assessment and Mitigation in Synthetic Tabular Data Generation", he evaluates six synthetic data generators and shows they systematically increase classifier unfairness — an effect standard mitigation corrects on high-utility data.

    Fairness Synthetic data Mitigation
  • 10 sep Defense

    Rodrigo Pagliusi defends a master's dissertation on fairness

    In the dissertation "Evaluation of Machine-Learning Classifiers Under Progressive Unfairness", he proposes Systematic Label Flipping for Fairness Stress Testing — controlled data bias to gradually measure how much traditional classifiers degrade under growing unfairness.

    Fairness Stress testing COMPAS
2024 archive

August 2024

01 item
  • 22 aug Defense

    Ygor Canalli defends a PhD on fairness in machine learning

    In the doctoral thesis "Robust Loss and Penalty for Fair Machine Learning", he explores the convergence of fairness and label noise and proposes Fair Transition Loss and Redlining Penalty Regularization, outperforming classical and state-of-the-art approaches on most fair-classification benchmarks.

    Fairness Robustness Label noise