Ethical Machine Learning
summary
Real examples: a parole denial, Amazon's hiring bias, UK exam downgrades and a model finding otherwise-undetectable breast cancer, plus deployed edge cameras for home alerts. Covers where bias appears — data, algorithm, model drift and actions — and concrete tooling: SageMaker Clarify for pre-training checks, reports with metrics like class imbalance 0.557 and DPL 0.018. Explains remediation steps (resample, impute, monitor, human review), legal constraints (GDPR, US AI Bill of Rights) and acceptable accuracy tradeoffs.
AI-written summary · about
chapters (5)
- 0:00Start
- 6:10Machine Learning is Widespread
- 9:16Traditional Software Development
- 22:02The outcome of biased decisions
- 23:52Bias shows up in your
related talks
54mEthical Machine Learning
55mEthical Machine Learning
47mEthical Machine Learning
23mMachine Learning Ethics
37mKeynote: Ethical Machine Learning: Creating Fair Models in an Unjust World
41mMachine Learning Model Fairness and Interpretability with AWS
47mEthics & AI: Identifying & preventing bias in predictive models
35mBuilding Inclusive Machine Learning Models