MC 5501
Speaker
Maria Gavrilova, PhD, University of Calgary
Title
Fair Play: Spotting Data and Machine Learning Bias in Real-World Datasets
Abstract
Recent advancements in artificial intelligence (AI) have raised significant ethical and philosophical challenges, particularly related to unequal access to healthcare. Progress has been made in understanding the impact of AI on health inequality, such as technology-based studies aiming to define and mitigate algorithmic bias and evaluate fairness in healthcare. However, health inequality remains a complex issue. In this lecture, we describe a fusion-based methodology for bias mitigation in medical data. Furthermore, we explore the relationships between demographic and clinical features and model bias, thereby adding an explainability dimension to the proposed bias mitigation strategies. Considering direct real-world applications allows us to address the complex issues of healthcare inequality, a problem that transcends across different demographics and disproportionally impacts marginalized communities.
This lecture is presented jointly by the Department of Applied Mathematics and the Women in Mathematics Committee