Public online seminar, 9am 6 August 2020 AEST
Clinton Castro of Florida International University will give the fourth HMI Data, AI and Society public seminar.
Clinton Castro is an assistant professor in the Department of Philosophy at Florida International University in Miami, Florida. His primary areas of study are epistemology (including formal) and ethics (especially information/data ethics). His current interests include fairness in machine learning and the ethics of predictive policing.
Clinton will discuss his work 'Just Machines'.
In this talk, Clinton will discuss how a number of recent findings in the field of machine learning have given rise to questions about what it means for automated scoring- or decision-making systems to be fair. One center of gravity in this discussion is whether such systems ought to satisfy classification parity (which requires parity in accuracy across groups defined by protected attributes) or calibration (which requires similar predictions to have similar meanings across groups defined by protected attributes). Central to this discussion are impossibility results, owed to Kleinberg et al.(2016), Chouldechova (2017), and Corbett-Davies et al. (2017), which show that classification parity and calibration are often incompatible. This paper aims to do three things: expose more philosophers to key concepts and ideas in the study of fair machine learning, argue that both classification parity and calibration are unsatisfactory measures of fairness, and offer a general diagnosis of the failure of these popular measures.