Paradoxes in Fair Machine Learning

            Paul Glz, Anson Kahng, and Ariel D. Procaccia
                 Computer Science Department
                  Carnegie Mellon University
             {pgoelz, akahng, arielpro}@cs.cmu.edu

                       Abstract
     Equalized odds is a statistical notion of fairness in machine learning that ensures
     that classification algorithms do not discriminate against protected groups. We
     extend equalized odds to the setting of cardinality- ...