Ensuring Fairness and Transparency in Artificial Intelligence: Best Practices and Techniques

Funda Güneş, Director of the Master's Program, Statistical Science, Duke University

This talk delves into the important topic of fairness and accountability in artificial intelligence systems. Attendees will learn about the various ways bias can creep into AI and how to quantify the extent of it. Techniques such as model debugging and error analysis, post-hoc model explanations, interpretability and causality, and human-in-the-loop approaches will be discussed as ways to improve the accountability of models and make AI more transparent and interpretable.

Bio: Funda Güneş is the Director of the Master's Program in Statistical Science at Duke University. With over a decade of experience as a machine learning researcher at SAS Institute, she brings a wealth of knowledge and expertise to her students, focusing her research on combining well-established statistical methods with algorithmic approaches to solve challenging predictive modeling problems, particularly in the area of health equity and responsible AI, and holds a patent for an efficient automated machine learning technique. Dr. Güneş is dedicated to mentoring and advising students and supporting women in the data science community, creating numerous learning and networking opportunities through partnerships with leading companies and accomplished professionals to foster a supportive and inclusive environment for all.