Algorithmic Fairness

concepts · 13 notes linked

Related: AI Ethics · AI Bias · Google · Timnit Gebru · AI Safety · Mit · Machine Learning · Bias In AI

Notes

• 160k+ high school students will only graduate if a statistical model allows them to — Critique of IB using statistical model to assign high-stakes graduation grades
• A new way to evaluate the impact of medical research — Diversity factor metric for medical research beyond citation impact factor
• ACM FAccT — Interdisciplinary conference on fairness accountability transparency in algorithmic systems
• AIF360/aif360/algorithms/inprocessing/adversarial_debiasing.py at main · Trusted-AI/AIF360 — AIF360 in-processing adversarial debiasing classifier implementation in TensorFlow
• AIF360/aif360/algorithms/inprocessing/adversarial_debiasing.py at main · Trusted-AI/AIF360 — In-processing adversarial debiasing classifier for fairness
• Bias in Criminal Risk Scores Is Mathematically Inevitable, Researchers Say — Mathematical proof that predictive parity and equal error rates are incompatible fairness criteria
• GitHub - daviddao/awful-ai: Awful AI is a curated list to track current scary usages of AI - hoping to raise awareness — Curated list cataloguing harmful and unethical real-world AI deployments
• Google says it's committed to ethical AI research. Its ethical AI team isn't so sure. — Google ethical AI team dysfunction after Timnit Gebru firing
• How machine-learning models can amplify inequities in medical diagnosis and treatment — Four-type taxonomy of ML subpopulation shifts causing medical inequity
• Justice by the Numbers: Meet the Statistician Trying to Fix Bias in Criminal Justice Algorithms — Kristian Lum using statistics to expose and fix bias in criminal justice algorithms
• Participation-washing could be the next dangerous fad in machine learning — Critique of performative participation as insufficient fix for ML injustice
• The Kidney Transplant Algorithm's Surprising Lessons for Ethical A.I. — Kidney allocation algorithm as model for inclusive AI governance
• Unveiling the Hidden Biases in Medical AI - Paving the Way for Fairer and More Accurate Imaging Diagnoses — 29 identified bias sources across the medical imaging AI development pipeline