A Gateway to Trustworthy AI: Using Visual Analytics to Unmask Coincidental Correlations
Explore the intersection of visual analytics and trustworthy AI in this hour-long lecture presented by Yu-Ru Lin from the University of Pittsburgh. Delve into innovative approaches for addressing challenges in machine learning and data-driven decision-making, focusing on the risks of spurious and biased associations. Learn about human-in-the-loop workflows designed to tackle AI blindspots in classification models, offering visually interpretable statistical methods to quantify and understand concept associations. Discover debiasing techniques to address misleading patterns in data. Examine the phenomenon of Simpson's Paradox and its impact on data interpretation, and explore an intuitive causal analysis framework that enables users to identify, understand, and prevent spurious associations. Gain insights into making AI more trustworthy through advanced visual analytics, contributing to more accountable causal decision-making in the field of artificial intelligence.