Algorithmic Fairness
Algorithmic Fairness is a research project that aims at mitigating any form of discrimination that might be created, perpetuated or increased when a computational system is designed -- espetially artificial intelligence systems. This project goal is to cover many aspects in fairness: (1) Identification of applications that might increase or propagate bias against minorities; (2) Design experiments/applications that investigate the fairness of already established machine learning models, particularly on application in Latin America; (3) Proposition of algorithms that are flexible enough to reduce fairness; (4) Create systems that helps experts to select the most appropriate model on each application, balancing fairness and accuracy goals.