Upper School Intersession 2025

Thanks to significant advances in machine learning and data science, decision makers are embracing and employing advanced algorithms and statistical models to help with or fully automate difficult tasks across our society. With examples ranging from advertising and finance to healthcare and criminal justice, machine learning tools have become ubiquitous. While often providing significant improvements in speed and performance, these tools come with increased complexity that can make the decision making process opaque and difficult to evaluate. How did your model make that prediction? Why? Are the decisions that it makes fair? How can we quantify fairness? In this activity, we will discuss real-world examples of automated algorithmic decision making along with the practical and ethical problems they can face. We will explore the ideas of bias, fairness, safety, and interpretability.