Fairness in Machine Learning: Tasks, Experiences, and Performance Measures

Explore the complexities of fairness in machine learning through this insightful lecture by Tina Eliassi-Rad from Northeastern University. Delve into the challenges of defining and implementing fairness in AI systems, examining current tasks, experiences, and performance measures. Analyze the limitations of risk assessment models and their impact on decision-making processes. Discover alternative approaches to providing context for human decision-makers. Investigate the "under-sampled majority" problem and its consequences on AI performance for diverse populations. Examine various fairness definitions, from group to individual fairness, and learn about a null model for measuring favoritism and prejudice in data. Gain valuable insights into the ethical considerations and practical challenges of developing fair and unbiased machine learning algorithms.

Just Machine Learning