How F.A.T. is Your ML Model Quality in the Era of Software

Explore a comprehensive approach to evaluating machine learning model quality beyond accuracy in this 45-minute conference talk by Yiannis Kanellopoulos, founder of Code4Thought. Delve into the critical F.A.T. properties - Fairness, Accountability, and Transparency - and their importance in responsible AI governance. Learn how to assess ML models using both qualitative and quantitative methods, including predefined checklists for technical and organizational governance, model-agnostic explanation mechanisms for post-hoc insights, and class-sensitive error rate metrics for bias testing. Gain valuable insights from real-world case studies demonstrating the benefits of making ML models accountable, transparent, and fair in the era of software development.