Explore a thought-provoking conference talk that delves into the intersection of feature importance methods in machine learning and feminist epistemology. Discover how epistemic values shape the development and interpretation of these methods, drawing valuable insights from feminist philosophy of science. Gain a deeper understanding of the underlying assumptions and biases in feature importance techniques, and learn how to critically evaluate their implications for fairness and accountability in AI systems. This 17-minute presentation, delivered at the ACM FAccT 2021 conference, challenges conventional approaches and offers fresh perspectives on improving the ethical and epistemological foundations of machine learning interpretability.
Epistemic Values in Feature Importance Methods - Lessons From Feminist Epistemology
Association for Computing Machinery (ACM) via YouTube
Overview
Syllabus
Epistemic Values in Feature Importance Methods: Lessons From Feminist Epistemology
Taught by
ACM FAccT Conference