On Selection Bias and Fairness Issues in Machine Learning

Explore selection bias and fairness challenges in machine learning through this academic seminar delivered by Professor Stephan Clémençon from Telecom Paris, Institut Polytechnique de Paris. Delve into the critical issues surrounding how selection bias affects machine learning models and examine the complex relationship between bias and fairness in algorithmic decision-making. Learn about the theoretical foundations and practical implications of these challenges, understanding how selection mechanisms can introduce systematic errors that compromise model performance and fairness across different populations. Discover methodological approaches for identifying, measuring, and mitigating selection bias while maintaining fairness principles in machine learning applications. Gain insights into the mathematical frameworks used to analyze these phenomena and explore real-world scenarios where selection bias and fairness concerns intersect. This presentation is part of the "Representing, calibrating & leveraging prediction uncertainty from statistics to machine learning" research programme at the Isaac Newton Institute for Mathematical Sciences.