Discrimination by AI is widespread, resulting in some groups being treated unfairly by systems that incorporate AI. The field of algorithmic fairness studies methods for combating algorithmic discrimination. An essential step is formalizing a notion of fairness for classifiers. I will demonstrate why this is more challenging than one might initially assume, and discuss types of formal fairness notions. I will then consider the challenge of auditing the fairness of classifiers that are not directly accessible, such as proprietary classifiers used by private companies. This requires a principled approach for quantifying unfairness, as well as methods for drawing conclusions on classifiers from limited aggregate statistics. I will present recent advances that show how we can prove that a classifier is unfair using limited information.
This event is open to the public with emphasis on graduate students in machine learning, computer science, ECE, statistics, mathematics, linguistics, medicine, as well as PhD-level data scientists in the GTA.