Responsible AI

Privacy, fairness, accountability and honest reporting, with code where code helps: re-identification risk, federated learning and measured rather than claimed results.

This part builds on Machine learning.

0 of 4 topics ready, listed in reading order

Privacy and anonymizationWhy removing names is not anonymization, k-anonymity and differential privacy.Planned

Federated learningTraining across devices without collecting their data.Planned

Fairness and accountabilityBias, transparency and who answers for a model's decisions.Planned

Honest evaluationReproducible experiments and reporting only what was measured.Planned