CITP Seminar: Kristian Lum - Closer Than They Appear: A Bayesian Perspective on Individual-level Heterogeneity in Risk Assessment
We find that individuals within the same risk group vary widely in their probability of the outcome. In practice, this means that allocating individuals to risk groups based on standard approaches to risk assessment, in large part, results in creating distinctions among individuals who are not meaningfully different in terms of their likelihood of the outcome. This is because uncertainty about the probability that any particular individual will fail to appear is large relative to the difference in average probabilities among any reasonable set of risk groups.
Kristian Lum is a senior staff machine learning researcher at Twitter in the Machine Learning Ethics, Transparency, and Accountability group. Prior to that she was a research professor at the University of Pennsylvania in the Department of Computer and Information Science and the lead statistician at the Human Rights Data Analysis Group. She was a founding member of the Executive Committee of the ACM Conference on Fairness, Accountability, and Transparency. Her research focuses on the responsible use of algorithmic decision-making, with an emphasis on evaluation of models for harmful impacts and mitigation techniques. In the past, her research has focused on the (un)fairness of predictive models used in the criminal justice system.